{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.01,
  "eval_steps": 500,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1e-05,
      "grad_norm": 1.474351326504247,
      "learning_rate": 3e-06,
      "loss": 10.8527,
      "step": 1
    },
    {
      "epoch": 2e-05,
      "grad_norm": 1.4679683222098754,
      "learning_rate": 6e-06,
      "loss": 10.8518,
      "step": 2
    },
    {
      "epoch": 3e-05,
      "grad_norm": 1.4778785546190916,
      "learning_rate": 9e-06,
      "loss": 10.8528,
      "step": 3
    },
    {
      "epoch": 4e-05,
      "grad_norm": 1.4606433182728087,
      "learning_rate": 1.2e-05,
      "loss": 10.8514,
      "step": 4
    },
    {
      "epoch": 5e-05,
      "grad_norm": 1.495032790614139,
      "learning_rate": 1.5e-05,
      "loss": 10.8475,
      "step": 5
    },
    {
      "epoch": 6e-05,
      "grad_norm": 1.4922781628880415,
      "learning_rate": 1.8e-05,
      "loss": 10.8456,
      "step": 6
    },
    {
      "epoch": 7e-05,
      "grad_norm": 1.4624391310631026,
      "learning_rate": 2.1000000000000002e-05,
      "loss": 10.8332,
      "step": 7
    },
    {
      "epoch": 8e-05,
      "grad_norm": 1.3590231342483396,
      "learning_rate": 2.4e-05,
      "loss": 10.8048,
      "step": 8
    },
    {
      "epoch": 9e-05,
      "grad_norm": 1.3125465506408707,
      "learning_rate": 2.7e-05,
      "loss": 10.7995,
      "step": 9
    },
    {
      "epoch": 0.0001,
      "grad_norm": 1.2786530848646291,
      "learning_rate": 3e-05,
      "loss": 10.7845,
      "step": 10
    },
    {
      "epoch": 0.00011,
      "grad_norm": 1.173935963529719,
      "learning_rate": 3.2999999999999996e-05,
      "loss": 10.7675,
      "step": 11
    },
    {
      "epoch": 0.00012,
      "grad_norm": 1.1421003663969806,
      "learning_rate": 3.6e-05,
      "loss": 10.7555,
      "step": 12
    },
    {
      "epoch": 0.00013,
      "grad_norm": 1.0874053290178312,
      "learning_rate": 3.9e-05,
      "loss": 10.7368,
      "step": 13
    },
    {
      "epoch": 0.00014,
      "grad_norm": 1.0659709149955685,
      "learning_rate": 4.2000000000000004e-05,
      "loss": 10.725,
      "step": 14
    },
    {
      "epoch": 0.00015,
      "grad_norm": 1.0504212767689105,
      "learning_rate": 4.4999999999999996e-05,
      "loss": 10.7164,
      "step": 15
    },
    {
      "epoch": 0.00016,
      "grad_norm": 1.0094105445485244,
      "learning_rate": 4.8e-05,
      "loss": 10.6992,
      "step": 16
    },
    {
      "epoch": 0.00017,
      "grad_norm": 0.9821324312557077,
      "learning_rate": 5.1000000000000006e-05,
      "loss": 10.6847,
      "step": 17
    },
    {
      "epoch": 0.00018,
      "grad_norm": 0.9678398856931318,
      "learning_rate": 5.4e-05,
      "loss": 10.6682,
      "step": 18
    },
    {
      "epoch": 0.00019,
      "grad_norm": 0.9461041060244328,
      "learning_rate": 5.7e-05,
      "loss": 10.6548,
      "step": 19
    },
    {
      "epoch": 0.0002,
      "grad_norm": 0.9477034430151614,
      "learning_rate": 6e-05,
      "loss": 10.643,
      "step": 20
    },
    {
      "epoch": 0.00021,
      "grad_norm": 0.9315183579373711,
      "learning_rate": 6.3e-05,
      "loss": 10.6292,
      "step": 21
    },
    {
      "epoch": 0.00022,
      "grad_norm": 0.9306702069422275,
      "learning_rate": 6.599999999999999e-05,
      "loss": 10.6138,
      "step": 22
    },
    {
      "epoch": 0.00023,
      "grad_norm": 0.9273403280850633,
      "learning_rate": 6.9e-05,
      "loss": 10.6009,
      "step": 23
    },
    {
      "epoch": 0.00024,
      "grad_norm": 0.9220363741136194,
      "learning_rate": 7.2e-05,
      "loss": 10.5875,
      "step": 24
    },
    {
      "epoch": 0.00025,
      "grad_norm": 0.9201767792857789,
      "learning_rate": 7.500000000000001e-05,
      "loss": 10.5733,
      "step": 25
    },
    {
      "epoch": 0.00026,
      "grad_norm": 0.9175580667413662,
      "learning_rate": 7.8e-05,
      "loss": 10.5584,
      "step": 26
    },
    {
      "epoch": 0.00027,
      "grad_norm": 0.9079543629096309,
      "learning_rate": 8.1e-05,
      "loss": 10.5463,
      "step": 27
    },
    {
      "epoch": 0.00028,
      "grad_norm": 0.9048806062272016,
      "learning_rate": 8.400000000000001e-05,
      "loss": 10.5323,
      "step": 28
    },
    {
      "epoch": 0.00029,
      "grad_norm": 0.9080906603458408,
      "learning_rate": 8.7e-05,
      "loss": 10.5179,
      "step": 29
    },
    {
      "epoch": 0.0003,
      "grad_norm": 0.9143257896829334,
      "learning_rate": 8.999999999999999e-05,
      "loss": 10.5023,
      "step": 30
    },
    {
      "epoch": 0.00031,
      "grad_norm": 0.9133459401307438,
      "learning_rate": 9.3e-05,
      "loss": 10.4873,
      "step": 31
    },
    {
      "epoch": 0.00032,
      "grad_norm": 0.9088988454650742,
      "learning_rate": 9.6e-05,
      "loss": 10.4732,
      "step": 32
    },
    {
      "epoch": 0.00033,
      "grad_norm": 0.9091480228096109,
      "learning_rate": 9.900000000000001e-05,
      "loss": 10.4568,
      "step": 33
    },
    {
      "epoch": 0.00034,
      "grad_norm": 0.9093347160978491,
      "learning_rate": 0.00010200000000000001,
      "loss": 10.4401,
      "step": 34
    },
    {
      "epoch": 0.00035,
      "grad_norm": 0.9151059868353938,
      "learning_rate": 0.00010500000000000002,
      "loss": 10.4221,
      "step": 35
    },
    {
      "epoch": 0.00036,
      "grad_norm": 0.9109062872419206,
      "learning_rate": 0.000108,
      "loss": 10.4047,
      "step": 36
    },
    {
      "epoch": 0.00037,
      "grad_norm": 0.9005574122536919,
      "learning_rate": 0.000111,
      "loss": 10.3878,
      "step": 37
    },
    {
      "epoch": 0.00038,
      "grad_norm": 0.9059566976246669,
      "learning_rate": 0.000114,
      "loss": 10.3677,
      "step": 38
    },
    {
      "epoch": 0.00039,
      "grad_norm": 0.909004498745371,
      "learning_rate": 0.000117,
      "loss": 10.3483,
      "step": 39
    },
    {
      "epoch": 0.0004,
      "grad_norm": 0.9133826925572958,
      "learning_rate": 0.00012,
      "loss": 10.3258,
      "step": 40
    },
    {
      "epoch": 0.00041,
      "grad_norm": 0.9039563412032365,
      "learning_rate": 0.000123,
      "loss": 10.3065,
      "step": 41
    },
    {
      "epoch": 0.00042,
      "grad_norm": 0.9101200951645674,
      "learning_rate": 0.000126,
      "loss": 10.2853,
      "step": 42
    },
    {
      "epoch": 0.00043,
      "grad_norm": 0.9146848302450299,
      "learning_rate": 0.000129,
      "loss": 10.2609,
      "step": 43
    },
    {
      "epoch": 0.00044,
      "grad_norm": 0.9151886453844831,
      "learning_rate": 0.00013199999999999998,
      "loss": 10.2386,
      "step": 44
    },
    {
      "epoch": 0.00045,
      "grad_norm": 0.9168774496112169,
      "learning_rate": 0.000135,
      "loss": 10.2149,
      "step": 45
    },
    {
      "epoch": 0.00046,
      "grad_norm": 0.9075399781785387,
      "learning_rate": 0.000138,
      "loss": 10.1934,
      "step": 46
    },
    {
      "epoch": 0.00047,
      "grad_norm": 0.9135459535669499,
      "learning_rate": 0.000141,
      "loss": 10.1656,
      "step": 47
    },
    {
      "epoch": 0.00048,
      "grad_norm": 0.91262035081973,
      "learning_rate": 0.000144,
      "loss": 10.141,
      "step": 48
    },
    {
      "epoch": 0.00049,
      "grad_norm": 0.9110594218495042,
      "learning_rate": 0.000147,
      "loss": 10.1163,
      "step": 49
    },
    {
      "epoch": 0.0005,
      "grad_norm": 0.9132890933284591,
      "learning_rate": 0.00015000000000000001,
      "loss": 10.0897,
      "step": 50
    },
    {
      "epoch": 0.00051,
      "grad_norm": 0.9150426988347248,
      "learning_rate": 0.000153,
      "loss": 10.0624,
      "step": 51
    },
    {
      "epoch": 0.00052,
      "grad_norm": 0.9109744535674626,
      "learning_rate": 0.000156,
      "loss": 10.0358,
      "step": 52
    },
    {
      "epoch": 0.00053,
      "grad_norm": 0.901532060032359,
      "learning_rate": 0.000159,
      "loss": 10.0114,
      "step": 53
    },
    {
      "epoch": 0.00054,
      "grad_norm": 0.9209685410544358,
      "learning_rate": 0.000162,
      "loss": 9.977,
      "step": 54
    },
    {
      "epoch": 0.00055,
      "grad_norm": 0.9128557250148688,
      "learning_rate": 0.000165,
      "loss": 9.9526,
      "step": 55
    },
    {
      "epoch": 0.00056,
      "grad_norm": 0.9040444969022912,
      "learning_rate": 0.00016800000000000002,
      "loss": 9.9235,
      "step": 56
    },
    {
      "epoch": 0.00057,
      "grad_norm": 0.9096304515854158,
      "learning_rate": 0.000171,
      "loss": 9.8951,
      "step": 57
    },
    {
      "epoch": 0.00058,
      "grad_norm": 0.9067337485581307,
      "learning_rate": 0.000174,
      "loss": 9.869,
      "step": 58
    },
    {
      "epoch": 0.00059,
      "grad_norm": 0.9110609200320162,
      "learning_rate": 0.000177,
      "loss": 9.8368,
      "step": 59
    },
    {
      "epoch": 0.0006,
      "grad_norm": 0.898536280212388,
      "learning_rate": 0.00017999999999999998,
      "loss": 9.8104,
      "step": 60
    },
    {
      "epoch": 0.00061,
      "grad_norm": 0.8973231498486932,
      "learning_rate": 0.000183,
      "loss": 9.781,
      "step": 61
    },
    {
      "epoch": 0.00062,
      "grad_norm": 0.9031035316544737,
      "learning_rate": 0.000186,
      "loss": 9.749,
      "step": 62
    },
    {
      "epoch": 0.00063,
      "grad_norm": 0.895903440256181,
      "learning_rate": 0.000189,
      "loss": 9.7201,
      "step": 63
    },
    {
      "epoch": 0.00064,
      "grad_norm": 0.892797539460924,
      "learning_rate": 0.000192,
      "loss": 9.6896,
      "step": 64
    },
    {
      "epoch": 0.00065,
      "grad_norm": 0.9002961211110138,
      "learning_rate": 0.00019500000000000002,
      "loss": 9.6554,
      "step": 65
    },
    {
      "epoch": 0.00066,
      "grad_norm": 0.8937380482279099,
      "learning_rate": 0.00019800000000000002,
      "loss": 9.6306,
      "step": 66
    },
    {
      "epoch": 0.00067,
      "grad_norm": 0.8953750022876055,
      "learning_rate": 0.000201,
      "loss": 9.6003,
      "step": 67
    },
    {
      "epoch": 0.00068,
      "grad_norm": 0.9015956584917496,
      "learning_rate": 0.00020400000000000003,
      "loss": 9.5664,
      "step": 68
    },
    {
      "epoch": 0.00069,
      "grad_norm": 0.8870475418543912,
      "learning_rate": 0.00020700000000000002,
      "loss": 9.5382,
      "step": 69
    },
    {
      "epoch": 0.0007,
      "grad_norm": 0.9005981766349919,
      "learning_rate": 0.00021000000000000004,
      "loss": 9.4999,
      "step": 70
    },
    {
      "epoch": 0.00071,
      "grad_norm": 0.8908379991096952,
      "learning_rate": 0.00021299999999999997,
      "loss": 9.4723,
      "step": 71
    },
    {
      "epoch": 0.00072,
      "grad_norm": 0.89454119786981,
      "learning_rate": 0.000216,
      "loss": 9.4447,
      "step": 72
    },
    {
      "epoch": 0.00073,
      "grad_norm": 0.8950328287787978,
      "learning_rate": 0.00021899999999999998,
      "loss": 9.4098,
      "step": 73
    },
    {
      "epoch": 0.00074,
      "grad_norm": 0.8955182261448745,
      "learning_rate": 0.000222,
      "loss": 9.3823,
      "step": 74
    },
    {
      "epoch": 0.00075,
      "grad_norm": 0.8910449269411236,
      "learning_rate": 0.000225,
      "loss": 9.3459,
      "step": 75
    },
    {
      "epoch": 0.00076,
      "grad_norm": 0.8994830990969283,
      "learning_rate": 0.000228,
      "loss": 9.3173,
      "step": 76
    },
    {
      "epoch": 0.00077,
      "grad_norm": 0.8945465759020658,
      "learning_rate": 0.000231,
      "loss": 9.2874,
      "step": 77
    },
    {
      "epoch": 0.00078,
      "grad_norm": 0.8885636678908063,
      "learning_rate": 0.000234,
      "loss": 9.2563,
      "step": 78
    },
    {
      "epoch": 0.00079,
      "grad_norm": 0.8922899375029398,
      "learning_rate": 0.00023700000000000001,
      "loss": 9.2188,
      "step": 79
    },
    {
      "epoch": 0.0008,
      "grad_norm": 0.8928864834710244,
      "learning_rate": 0.00024,
      "loss": 9.1835,
      "step": 80
    },
    {
      "epoch": 0.00081,
      "grad_norm": 0.898627735558165,
      "learning_rate": 0.00024300000000000002,
      "loss": 9.1509,
      "step": 81
    },
    {
      "epoch": 0.00082,
      "grad_norm": 0.9056278867565712,
      "learning_rate": 0.000246,
      "loss": 9.126,
      "step": 82
    },
    {
      "epoch": 0.00083,
      "grad_norm": 0.8959873225353988,
      "learning_rate": 0.00024900000000000004,
      "loss": 9.0897,
      "step": 83
    },
    {
      "epoch": 0.00084,
      "grad_norm": 0.8992390398731138,
      "learning_rate": 0.000252,
      "loss": 9.0582,
      "step": 84
    },
    {
      "epoch": 0.00085,
      "grad_norm": 0.8929295183813671,
      "learning_rate": 0.000255,
      "loss": 9.0339,
      "step": 85
    },
    {
      "epoch": 0.00086,
      "grad_norm": 0.895130193545426,
      "learning_rate": 0.000258,
      "loss": 8.9969,
      "step": 86
    },
    {
      "epoch": 0.00087,
      "grad_norm": 0.8878745708200014,
      "learning_rate": 0.000261,
      "loss": 8.9749,
      "step": 87
    },
    {
      "epoch": 0.00088,
      "grad_norm": 0.8916925738718505,
      "learning_rate": 0.00026399999999999997,
      "loss": 8.9377,
      "step": 88
    },
    {
      "epoch": 0.00089,
      "grad_norm": 0.8869518010064994,
      "learning_rate": 0.000267,
      "loss": 8.9112,
      "step": 89
    },
    {
      "epoch": 0.0009,
      "grad_norm": 0.8862477905139782,
      "learning_rate": 0.00027,
      "loss": 8.8834,
      "step": 90
    },
    {
      "epoch": 0.00091,
      "grad_norm": 0.8806789869277061,
      "learning_rate": 0.000273,
      "loss": 8.8534,
      "step": 91
    },
    {
      "epoch": 0.00092,
      "grad_norm": 0.8827535811867636,
      "learning_rate": 0.000276,
      "loss": 8.8203,
      "step": 92
    },
    {
      "epoch": 0.00093,
      "grad_norm": 0.8863633123718828,
      "learning_rate": 0.000279,
      "loss": 8.7906,
      "step": 93
    },
    {
      "epoch": 0.00094,
      "grad_norm": 0.8854377864602295,
      "learning_rate": 0.000282,
      "loss": 8.7643,
      "step": 94
    },
    {
      "epoch": 0.00095,
      "grad_norm": 0.8811122089944353,
      "learning_rate": 0.000285,
      "loss": 8.7332,
      "step": 95
    },
    {
      "epoch": 0.00096,
      "grad_norm": 0.8875459668079012,
      "learning_rate": 0.000288,
      "loss": 8.701,
      "step": 96
    },
    {
      "epoch": 0.00097,
      "grad_norm": 0.8790489647826922,
      "learning_rate": 0.000291,
      "loss": 8.6763,
      "step": 97
    },
    {
      "epoch": 0.00098,
      "grad_norm": 0.8816372646189595,
      "learning_rate": 0.000294,
      "loss": 8.6393,
      "step": 98
    },
    {
      "epoch": 0.00099,
      "grad_norm": 0.8790820085994937,
      "learning_rate": 0.000297,
      "loss": 8.6187,
      "step": 99
    },
    {
      "epoch": 0.001,
      "grad_norm": 0.8752174625291169,
      "learning_rate": 0.00030000000000000003,
      "loss": 8.5928,
      "step": 100
    },
    {
      "epoch": 0.00101,
      "grad_norm": 0.8773415713400505,
      "learning_rate": 0.00030300000000000005,
      "loss": 8.5584,
      "step": 101
    },
    {
      "epoch": 0.00102,
      "grad_norm": 0.8643663486077326,
      "learning_rate": 0.000306,
      "loss": 8.5391,
      "step": 102
    },
    {
      "epoch": 0.00103,
      "grad_norm": 0.8697458804733939,
      "learning_rate": 0.000309,
      "loss": 8.5142,
      "step": 103
    },
    {
      "epoch": 0.00104,
      "grad_norm": 0.8606206924682385,
      "learning_rate": 0.000312,
      "loss": 8.4945,
      "step": 104
    },
    {
      "epoch": 0.00105,
      "grad_norm": 0.8647215352159486,
      "learning_rate": 0.000315,
      "loss": 8.4678,
      "step": 105
    },
    {
      "epoch": 0.00106,
      "grad_norm": 0.8616489218188034,
      "learning_rate": 0.000318,
      "loss": 8.4385,
      "step": 106
    },
    {
      "epoch": 0.00107,
      "grad_norm": 0.8529010606970446,
      "learning_rate": 0.000321,
      "loss": 8.4145,
      "step": 107
    },
    {
      "epoch": 0.00108,
      "grad_norm": 0.8573365553662458,
      "learning_rate": 0.000324,
      "loss": 8.3854,
      "step": 108
    },
    {
      "epoch": 0.00109,
      "grad_norm": 0.8466244623266435,
      "learning_rate": 0.000327,
      "loss": 8.3656,
      "step": 109
    },
    {
      "epoch": 0.0011,
      "grad_norm": 0.8493987341311698,
      "learning_rate": 0.00033,
      "loss": 8.3429,
      "step": 110
    },
    {
      "epoch": 0.00111,
      "grad_norm": 0.8418183624218302,
      "learning_rate": 0.000333,
      "loss": 8.3123,
      "step": 111
    },
    {
      "epoch": 0.00112,
      "grad_norm": 0.837428520328162,
      "learning_rate": 0.00033600000000000004,
      "loss": 8.2891,
      "step": 112
    },
    {
      "epoch": 0.00113,
      "grad_norm": 0.8530708116750574,
      "learning_rate": 0.000339,
      "loss": 8.2481,
      "step": 113
    },
    {
      "epoch": 0.00114,
      "grad_norm": 0.8531960306338435,
      "learning_rate": 0.000342,
      "loss": 8.2404,
      "step": 114
    },
    {
      "epoch": 0.00115,
      "grad_norm": 0.9106108649120399,
      "learning_rate": 0.00034500000000000004,
      "loss": 8.2161,
      "step": 115
    },
    {
      "epoch": 0.00116,
      "grad_norm": 0.9754034767309321,
      "learning_rate": 0.000348,
      "loss": 8.1959,
      "step": 116
    },
    {
      "epoch": 0.00117,
      "grad_norm": 0.8915473944545516,
      "learning_rate": 0.000351,
      "loss": 8.164,
      "step": 117
    },
    {
      "epoch": 0.00118,
      "grad_norm": 0.8220334698299727,
      "learning_rate": 0.000354,
      "loss": 8.146,
      "step": 118
    },
    {
      "epoch": 0.00119,
      "grad_norm": 0.8743176947954969,
      "learning_rate": 0.000357,
      "loss": 8.1361,
      "step": 119
    },
    {
      "epoch": 0.0012,
      "grad_norm": 0.8050422574835271,
      "learning_rate": 0.00035999999999999997,
      "loss": 8.1077,
      "step": 120
    },
    {
      "epoch": 0.00121,
      "grad_norm": 0.8378364817917794,
      "learning_rate": 0.000363,
      "loss": 8.0814,
      "step": 121
    },
    {
      "epoch": 0.00122,
      "grad_norm": 0.7934725047304613,
      "learning_rate": 0.000366,
      "loss": 8.0639,
      "step": 122
    },
    {
      "epoch": 0.00123,
      "grad_norm": 0.8001278864423256,
      "learning_rate": 0.000369,
      "loss": 8.0414,
      "step": 123
    },
    {
      "epoch": 0.00124,
      "grad_norm": 0.7891502944141346,
      "learning_rate": 0.000372,
      "loss": 8.0193,
      "step": 124
    },
    {
      "epoch": 0.00125,
      "grad_norm": 0.795452065739791,
      "learning_rate": 0.000375,
      "loss": 7.9938,
      "step": 125
    },
    {
      "epoch": 0.00126,
      "grad_norm": 0.7774797219919642,
      "learning_rate": 0.000378,
      "loss": 7.981,
      "step": 126
    },
    {
      "epoch": 0.00127,
      "grad_norm": 0.7551582047426328,
      "learning_rate": 0.000381,
      "loss": 7.9562,
      "step": 127
    },
    {
      "epoch": 0.00128,
      "grad_norm": 0.7770412292819664,
      "learning_rate": 0.000384,
      "loss": 7.9363,
      "step": 128
    },
    {
      "epoch": 0.00129,
      "grad_norm": 0.730245978508778,
      "learning_rate": 0.00038700000000000003,
      "loss": 7.9242,
      "step": 129
    },
    {
      "epoch": 0.0013,
      "grad_norm": 0.7581733929203945,
      "learning_rate": 0.00039000000000000005,
      "loss": 7.8975,
      "step": 130
    },
    {
      "epoch": 0.00131,
      "grad_norm": 0.7437048547743047,
      "learning_rate": 0.000393,
      "loss": 7.8642,
      "step": 131
    },
    {
      "epoch": 0.00132,
      "grad_norm": 0.7447956363319407,
      "learning_rate": 0.00039600000000000003,
      "loss": 7.8553,
      "step": 132
    },
    {
      "epoch": 0.00133,
      "grad_norm": 0.8219877092942749,
      "learning_rate": 0.00039900000000000005,
      "loss": 7.8359,
      "step": 133
    },
    {
      "epoch": 0.00134,
      "grad_norm": 1.2937247547771387,
      "learning_rate": 0.000402,
      "loss": 7.8316,
      "step": 134
    },
    {
      "epoch": 0.00135,
      "grad_norm": 1.173069005114359,
      "learning_rate": 0.00040500000000000003,
      "loss": 7.8093,
      "step": 135
    },
    {
      "epoch": 0.00136,
      "grad_norm": 0.7468130576157733,
      "learning_rate": 0.00040800000000000005,
      "loss": 7.7827,
      "step": 136
    },
    {
      "epoch": 0.00137,
      "grad_norm": 0.9811726952576713,
      "learning_rate": 0.000411,
      "loss": 7.7674,
      "step": 137
    },
    {
      "epoch": 0.00138,
      "grad_norm": 0.9462188302991427,
      "learning_rate": 0.00041400000000000003,
      "loss": 7.7401,
      "step": 138
    },
    {
      "epoch": 0.00139,
      "grad_norm": 0.8654346045111051,
      "learning_rate": 0.00041700000000000005,
      "loss": 7.7263,
      "step": 139
    },
    {
      "epoch": 0.0014,
      "grad_norm": 0.7113280167772615,
      "learning_rate": 0.00042000000000000007,
      "loss": 7.706,
      "step": 140
    },
    {
      "epoch": 0.00141,
      "grad_norm": 0.9535594232109135,
      "learning_rate": 0.000423,
      "loss": 7.6966,
      "step": 141
    },
    {
      "epoch": 0.00142,
      "grad_norm": 0.721845056784479,
      "learning_rate": 0.00042599999999999995,
      "loss": 7.6788,
      "step": 142
    },
    {
      "epoch": 0.00143,
      "grad_norm": 0.7614673366661663,
      "learning_rate": 0.00042899999999999997,
      "loss": 7.6517,
      "step": 143
    },
    {
      "epoch": 0.00144,
      "grad_norm": 0.6558320856409382,
      "learning_rate": 0.000432,
      "loss": 7.6295,
      "step": 144
    },
    {
      "epoch": 0.00145,
      "grad_norm": 0.7045333447763749,
      "learning_rate": 0.000435,
      "loss": 7.6259,
      "step": 145
    },
    {
      "epoch": 0.00146,
      "grad_norm": 0.6284166988938972,
      "learning_rate": 0.00043799999999999997,
      "loss": 7.585,
      "step": 146
    },
    {
      "epoch": 0.00147,
      "grad_norm": 0.6467600462746109,
      "learning_rate": 0.000441,
      "loss": 7.5791,
      "step": 147
    },
    {
      "epoch": 0.00148,
      "grad_norm": 0.6213738634604956,
      "learning_rate": 0.000444,
      "loss": 7.5641,
      "step": 148
    },
    {
      "epoch": 0.00149,
      "grad_norm": 0.5966591129804824,
      "learning_rate": 0.00044699999999999997,
      "loss": 7.5597,
      "step": 149
    },
    {
      "epoch": 0.0015,
      "grad_norm": 0.5636118960239422,
      "learning_rate": 0.00045,
      "loss": 7.5245,
      "step": 150
    },
    {
      "epoch": 0.00151,
      "grad_norm": 0.5909619206615454,
      "learning_rate": 0.000453,
      "loss": 7.5045,
      "step": 151
    },
    {
      "epoch": 0.00152,
      "grad_norm": 0.5383178931411042,
      "learning_rate": 0.000456,
      "loss": 7.497,
      "step": 152
    },
    {
      "epoch": 0.00153,
      "grad_norm": 0.5291626942545249,
      "learning_rate": 0.000459,
      "loss": 7.4853,
      "step": 153
    },
    {
      "epoch": 0.00154,
      "grad_norm": 0.5532016756899171,
      "learning_rate": 0.000462,
      "loss": 7.4637,
      "step": 154
    },
    {
      "epoch": 0.00155,
      "grad_norm": 0.5008942629560345,
      "learning_rate": 0.000465,
      "loss": 7.4483,
      "step": 155
    },
    {
      "epoch": 0.00156,
      "grad_norm": 0.5258951319062264,
      "learning_rate": 0.000468,
      "loss": 7.4224,
      "step": 156
    },
    {
      "epoch": 0.00157,
      "grad_norm": 0.4967086695788387,
      "learning_rate": 0.000471,
      "loss": 7.4172,
      "step": 157
    },
    {
      "epoch": 0.00158,
      "grad_norm": 0.48191095037320825,
      "learning_rate": 0.00047400000000000003,
      "loss": 7.3957,
      "step": 158
    },
    {
      "epoch": 0.00159,
      "grad_norm": 0.46801796765212106,
      "learning_rate": 0.000477,
      "loss": 7.3849,
      "step": 159
    },
    {
      "epoch": 0.0016,
      "grad_norm": 0.5255102038727893,
      "learning_rate": 0.00048,
      "loss": 7.3658,
      "step": 160
    },
    {
      "epoch": 0.00161,
      "grad_norm": 0.5299389188763052,
      "learning_rate": 0.00048300000000000003,
      "loss": 7.3428,
      "step": 161
    },
    {
      "epoch": 0.00162,
      "grad_norm": 0.5953669493848933,
      "learning_rate": 0.00048600000000000005,
      "loss": 7.3479,
      "step": 162
    },
    {
      "epoch": 0.00163,
      "grad_norm": 0.5335280557878616,
      "learning_rate": 0.0004890000000000001,
      "loss": 7.3286,
      "step": 163
    },
    {
      "epoch": 0.00164,
      "grad_norm": 0.4971558665793995,
      "learning_rate": 0.000492,
      "loss": 7.3284,
      "step": 164
    },
    {
      "epoch": 0.00165,
      "grad_norm": 0.45817405077400825,
      "learning_rate": 0.000495,
      "loss": 7.302,
      "step": 165
    },
    {
      "epoch": 0.00166,
      "grad_norm": 0.5803800594997708,
      "learning_rate": 0.0004980000000000001,
      "loss": 7.2782,
      "step": 166
    },
    {
      "epoch": 0.00167,
      "grad_norm": 0.6082011920361033,
      "learning_rate": 0.000501,
      "loss": 7.2708,
      "step": 167
    },
    {
      "epoch": 0.00168,
      "grad_norm": 0.6336998596864457,
      "learning_rate": 0.000504,
      "loss": 7.2488,
      "step": 168
    },
    {
      "epoch": 0.00169,
      "grad_norm": 0.6003268704217197,
      "learning_rate": 0.0005070000000000001,
      "loss": 7.2442,
      "step": 169
    },
    {
      "epoch": 0.0017,
      "grad_norm": 0.6942878050784774,
      "learning_rate": 0.00051,
      "loss": 7.2281,
      "step": 170
    },
    {
      "epoch": 0.00171,
      "grad_norm": 0.6828271953312538,
      "learning_rate": 0.000513,
      "loss": 7.2148,
      "step": 171
    },
    {
      "epoch": 0.00172,
      "grad_norm": 0.49790810289726045,
      "learning_rate": 0.000516,
      "loss": 7.2077,
      "step": 172
    },
    {
      "epoch": 0.00173,
      "grad_norm": 0.5828184415582448,
      "learning_rate": 0.0005189999999999999,
      "loss": 7.1934,
      "step": 173
    },
    {
      "epoch": 0.00174,
      "grad_norm": 0.5895089207006229,
      "learning_rate": 0.000522,
      "loss": 7.2103,
      "step": 174
    },
    {
      "epoch": 0.00175,
      "grad_norm": 0.7514233155115281,
      "learning_rate": 0.000525,
      "loss": 7.1773,
      "step": 175
    },
    {
      "epoch": 0.00176,
      "grad_norm": 0.4352152488230879,
      "learning_rate": 0.0005279999999999999,
      "loss": 7.144,
      "step": 176
    },
    {
      "epoch": 0.00177,
      "grad_norm": 0.4945639346042136,
      "learning_rate": 0.000531,
      "loss": 7.1524,
      "step": 177
    },
    {
      "epoch": 0.00178,
      "grad_norm": 0.4797513290452921,
      "learning_rate": 0.000534,
      "loss": 7.1405,
      "step": 178
    },
    {
      "epoch": 0.00179,
      "grad_norm": 0.4349662352142082,
      "learning_rate": 0.000537,
      "loss": 7.1325,
      "step": 179
    },
    {
      "epoch": 0.0018,
      "grad_norm": 0.38883631831142346,
      "learning_rate": 0.00054,
      "loss": 7.1029,
      "step": 180
    },
    {
      "epoch": 0.00181,
      "grad_norm": 0.49930253237147615,
      "learning_rate": 0.000543,
      "loss": 7.093,
      "step": 181
    },
    {
      "epoch": 0.00182,
      "grad_norm": 0.409986197327767,
      "learning_rate": 0.000546,
      "loss": 7.0882,
      "step": 182
    },
    {
      "epoch": 0.00183,
      "grad_norm": 0.4860898971704763,
      "learning_rate": 0.000549,
      "loss": 7.0781,
      "step": 183
    },
    {
      "epoch": 0.00184,
      "grad_norm": 0.5722795587528919,
      "learning_rate": 0.000552,
      "loss": 7.069,
      "step": 184
    },
    {
      "epoch": 0.00185,
      "grad_norm": 0.5738191189627982,
      "learning_rate": 0.000555,
      "loss": 7.0441,
      "step": 185
    },
    {
      "epoch": 0.00186,
      "grad_norm": 0.480936267251981,
      "learning_rate": 0.000558,
      "loss": 7.0527,
      "step": 186
    },
    {
      "epoch": 0.00187,
      "grad_norm": 0.6162211084806141,
      "learning_rate": 0.000561,
      "loss": 7.0307,
      "step": 187
    },
    {
      "epoch": 0.00188,
      "grad_norm": 0.5594962116008837,
      "learning_rate": 0.000564,
      "loss": 7.0081,
      "step": 188
    },
    {
      "epoch": 0.00189,
      "grad_norm": 0.5037393315861951,
      "learning_rate": 0.000567,
      "loss": 7.0028,
      "step": 189
    },
    {
      "epoch": 0.0019,
      "grad_norm": 0.4257079536133974,
      "learning_rate": 0.00057,
      "loss": 6.9836,
      "step": 190
    },
    {
      "epoch": 0.00191,
      "grad_norm": 0.4729880834795408,
      "learning_rate": 0.000573,
      "loss": 6.9775,
      "step": 191
    },
    {
      "epoch": 0.00192,
      "grad_norm": 0.35979347564719405,
      "learning_rate": 0.000576,
      "loss": 6.9672,
      "step": 192
    },
    {
      "epoch": 0.00193,
      "grad_norm": 0.3504590762022799,
      "learning_rate": 0.000579,
      "loss": 6.9673,
      "step": 193
    },
    {
      "epoch": 0.00194,
      "grad_norm": 0.40017651296086854,
      "learning_rate": 0.000582,
      "loss": 6.9599,
      "step": 194
    },
    {
      "epoch": 0.00195,
      "grad_norm": 0.4217845269426591,
      "learning_rate": 0.000585,
      "loss": 6.9408,
      "step": 195
    },
    {
      "epoch": 0.00196,
      "grad_norm": 0.3759474186154596,
      "learning_rate": 0.000588,
      "loss": 6.9274,
      "step": 196
    },
    {
      "epoch": 0.00197,
      "grad_norm": 0.38483250179950723,
      "learning_rate": 0.000591,
      "loss": 6.9235,
      "step": 197
    },
    {
      "epoch": 0.00198,
      "grad_norm": 0.5774981009451395,
      "learning_rate": 0.000594,
      "loss": 6.9171,
      "step": 198
    },
    {
      "epoch": 0.00199,
      "grad_norm": 0.7947838931607083,
      "learning_rate": 0.0005970000000000001,
      "loss": 6.905,
      "step": 199
    },
    {
      "epoch": 0.002,
      "grad_norm": 1.3095452790094997,
      "learning_rate": 0.0006000000000000001,
      "loss": 6.9171,
      "step": 200
    },
    {
      "epoch": 0.00201,
      "grad_norm": 0.8941477277896986,
      "learning_rate": 0.000603,
      "loss": 6.8832,
      "step": 201
    },
    {
      "epoch": 0.00202,
      "grad_norm": 0.6446211768569402,
      "learning_rate": 0.0006060000000000001,
      "loss": 6.8657,
      "step": 202
    },
    {
      "epoch": 0.00203,
      "grad_norm": 0.5529910396828944,
      "learning_rate": 0.0006090000000000001,
      "loss": 6.8657,
      "step": 203
    },
    {
      "epoch": 0.00204,
      "grad_norm": 0.6798364259433272,
      "learning_rate": 0.000612,
      "loss": 6.8577,
      "step": 204
    },
    {
      "epoch": 0.00205,
      "grad_norm": 0.4644570109539818,
      "learning_rate": 0.000615,
      "loss": 6.8377,
      "step": 205
    },
    {
      "epoch": 0.00206,
      "grad_norm": 0.44631259910699017,
      "learning_rate": 0.000618,
      "loss": 6.8297,
      "step": 206
    },
    {
      "epoch": 0.00207,
      "grad_norm": 0.5149457252986483,
      "learning_rate": 0.000621,
      "loss": 6.8239,
      "step": 207
    },
    {
      "epoch": 0.00208,
      "grad_norm": 0.38821898080366113,
      "learning_rate": 0.000624,
      "loss": 6.815,
      "step": 208
    },
    {
      "epoch": 0.00209,
      "grad_norm": 0.5320501684611929,
      "learning_rate": 0.000627,
      "loss": 6.8005,
      "step": 209
    },
    {
      "epoch": 0.0021,
      "grad_norm": 0.31085611137953467,
      "learning_rate": 0.00063,
      "loss": 6.7833,
      "step": 210
    },
    {
      "epoch": 0.00211,
      "grad_norm": 0.4810682865402516,
      "learning_rate": 0.000633,
      "loss": 6.7824,
      "step": 211
    },
    {
      "epoch": 0.00212,
      "grad_norm": 0.46892365964621546,
      "learning_rate": 0.000636,
      "loss": 6.7721,
      "step": 212
    },
    {
      "epoch": 0.00213,
      "grad_norm": 0.4643908337987038,
      "learning_rate": 0.000639,
      "loss": 6.7526,
      "step": 213
    },
    {
      "epoch": 0.00214,
      "grad_norm": 0.4868851924200075,
      "learning_rate": 0.000642,
      "loss": 6.7506,
      "step": 214
    },
    {
      "epoch": 0.00215,
      "grad_norm": 0.4395116450664113,
      "learning_rate": 0.000645,
      "loss": 6.7458,
      "step": 215
    },
    {
      "epoch": 0.00216,
      "grad_norm": 0.41354828342235656,
      "learning_rate": 0.000648,
      "loss": 6.74,
      "step": 216
    },
    {
      "epoch": 0.00217,
      "grad_norm": 0.3070152795385597,
      "learning_rate": 0.000651,
      "loss": 6.709,
      "step": 217
    },
    {
      "epoch": 0.00218,
      "grad_norm": 0.40442589086752323,
      "learning_rate": 0.000654,
      "loss": 6.7138,
      "step": 218
    },
    {
      "epoch": 0.00219,
      "grad_norm": 0.441644630454743,
      "learning_rate": 0.000657,
      "loss": 6.6872,
      "step": 219
    },
    {
      "epoch": 0.0022,
      "grad_norm": 0.6023964511244305,
      "learning_rate": 0.00066,
      "loss": 6.7066,
      "step": 220
    },
    {
      "epoch": 0.00221,
      "grad_norm": 0.7436617780357457,
      "learning_rate": 0.0006630000000000001,
      "loss": 6.6883,
      "step": 221
    },
    {
      "epoch": 0.00222,
      "grad_norm": 1.107124936944928,
      "learning_rate": 0.000666,
      "loss": 6.6857,
      "step": 222
    },
    {
      "epoch": 0.00223,
      "grad_norm": 0.9806174045781153,
      "learning_rate": 0.000669,
      "loss": 6.6846,
      "step": 223
    },
    {
      "epoch": 0.00224,
      "grad_norm": 0.5030190225010961,
      "learning_rate": 0.0006720000000000001,
      "loss": 6.6654,
      "step": 224
    },
    {
      "epoch": 0.00225,
      "grad_norm": 0.6877668848936607,
      "learning_rate": 0.000675,
      "loss": 6.6563,
      "step": 225
    },
    {
      "epoch": 0.00226,
      "grad_norm": 0.8124295484727102,
      "learning_rate": 0.000678,
      "loss": 6.655,
      "step": 226
    },
    {
      "epoch": 0.00227,
      "grad_norm": 0.7781712168651111,
      "learning_rate": 0.0006810000000000001,
      "loss": 6.6292,
      "step": 227
    },
    {
      "epoch": 0.00228,
      "grad_norm": 0.6536472724437511,
      "learning_rate": 0.000684,
      "loss": 6.6269,
      "step": 228
    },
    {
      "epoch": 0.00229,
      "grad_norm": 0.5457565622061897,
      "learning_rate": 0.000687,
      "loss": 6.6228,
      "step": 229
    },
    {
      "epoch": 0.0023,
      "grad_norm": 0.3628095747750561,
      "learning_rate": 0.0006900000000000001,
      "loss": 6.5952,
      "step": 230
    },
    {
      "epoch": 0.00231,
      "grad_norm": 0.5474995275109145,
      "learning_rate": 0.000693,
      "loss": 6.5849,
      "step": 231
    },
    {
      "epoch": 0.00232,
      "grad_norm": 0.40776460058160763,
      "learning_rate": 0.000696,
      "loss": 6.5921,
      "step": 232
    },
    {
      "epoch": 0.00233,
      "grad_norm": 0.35178189775679714,
      "learning_rate": 0.0006990000000000001,
      "loss": 6.5782,
      "step": 233
    },
    {
      "epoch": 0.00234,
      "grad_norm": 0.4281055336873759,
      "learning_rate": 0.000702,
      "loss": 6.5708,
      "step": 234
    },
    {
      "epoch": 0.00235,
      "grad_norm": 0.2803773137034946,
      "learning_rate": 0.000705,
      "loss": 6.5534,
      "step": 235
    },
    {
      "epoch": 0.00236,
      "grad_norm": 0.3921542773711977,
      "learning_rate": 0.000708,
      "loss": 6.5364,
      "step": 236
    },
    {
      "epoch": 0.00237,
      "grad_norm": 0.3526763779813035,
      "learning_rate": 0.0007109999999999999,
      "loss": 6.5448,
      "step": 237
    },
    {
      "epoch": 0.00238,
      "grad_norm": 0.5116945273918518,
      "learning_rate": 0.000714,
      "loss": 6.5357,
      "step": 238
    },
    {
      "epoch": 0.00239,
      "grad_norm": 0.6738279229074454,
      "learning_rate": 0.000717,
      "loss": 6.5224,
      "step": 239
    },
    {
      "epoch": 0.0024,
      "grad_norm": 0.8853347385085402,
      "learning_rate": 0.0007199999999999999,
      "loss": 6.5441,
      "step": 240
    },
    {
      "epoch": 0.00241,
      "grad_norm": 0.7775153220084405,
      "learning_rate": 0.000723,
      "loss": 6.5282,
      "step": 241
    },
    {
      "epoch": 0.00242,
      "grad_norm": 0.41795084065691346,
      "learning_rate": 0.000726,
      "loss": 6.4926,
      "step": 242
    },
    {
      "epoch": 0.00243,
      "grad_norm": 0.6402313725243838,
      "learning_rate": 0.000729,
      "loss": 6.4862,
      "step": 243
    },
    {
      "epoch": 0.00244,
      "grad_norm": 0.715394437807476,
      "learning_rate": 0.000732,
      "loss": 6.4949,
      "step": 244
    },
    {
      "epoch": 0.00245,
      "grad_norm": 0.8873416534047,
      "learning_rate": 0.000735,
      "loss": 6.487,
      "step": 245
    },
    {
      "epoch": 0.00246,
      "grad_norm": 1.0259114438292367,
      "learning_rate": 0.000738,
      "loss": 6.4771,
      "step": 246
    },
    {
      "epoch": 0.00247,
      "grad_norm": 0.6418814776118988,
      "learning_rate": 0.000741,
      "loss": 6.46,
      "step": 247
    },
    {
      "epoch": 0.00248,
      "grad_norm": 0.491074950554399,
      "learning_rate": 0.000744,
      "loss": 6.4659,
      "step": 248
    },
    {
      "epoch": 0.00249,
      "grad_norm": 0.5836449103944761,
      "learning_rate": 0.000747,
      "loss": 6.4457,
      "step": 249
    },
    {
      "epoch": 0.0025,
      "grad_norm": 0.4400008529473732,
      "learning_rate": 0.00075,
      "loss": 6.447,
      "step": 250
    },
    {
      "epoch": 0.00251,
      "grad_norm": 0.5056683711317587,
      "learning_rate": 0.000753,
      "loss": 6.4089,
      "step": 251
    },
    {
      "epoch": 0.00252,
      "grad_norm": 0.5475557445898553,
      "learning_rate": 0.000756,
      "loss": 6.4219,
      "step": 252
    },
    {
      "epoch": 0.00253,
      "grad_norm": 0.5310741382277282,
      "learning_rate": 0.000759,
      "loss": 6.41,
      "step": 253
    },
    {
      "epoch": 0.00254,
      "grad_norm": 0.5250408531352325,
      "learning_rate": 0.000762,
      "loss": 6.4057,
      "step": 254
    },
    {
      "epoch": 0.00255,
      "grad_norm": 0.3683514074420304,
      "learning_rate": 0.0007650000000000001,
      "loss": 6.4012,
      "step": 255
    },
    {
      "epoch": 0.00256,
      "grad_norm": 0.43783601316780707,
      "learning_rate": 0.000768,
      "loss": 6.3808,
      "step": 256
    },
    {
      "epoch": 0.00257,
      "grad_norm": 0.5249634850463766,
      "learning_rate": 0.000771,
      "loss": 6.3866,
      "step": 257
    },
    {
      "epoch": 0.00258,
      "grad_norm": 0.7257500993113621,
      "learning_rate": 0.0007740000000000001,
      "loss": 6.367,
      "step": 258
    },
    {
      "epoch": 0.00259,
      "grad_norm": 0.8020520964186684,
      "learning_rate": 0.000777,
      "loss": 6.3732,
      "step": 259
    },
    {
      "epoch": 0.0026,
      "grad_norm": 0.8803264650469431,
      "learning_rate": 0.0007800000000000001,
      "loss": 6.357,
      "step": 260
    },
    {
      "epoch": 0.00261,
      "grad_norm": 0.9480171721362259,
      "learning_rate": 0.0007830000000000001,
      "loss": 6.3728,
      "step": 261
    },
    {
      "epoch": 0.00262,
      "grad_norm": 0.8196081199065424,
      "learning_rate": 0.000786,
      "loss": 6.3478,
      "step": 262
    },
    {
      "epoch": 0.00263,
      "grad_norm": 0.5320550716287101,
      "learning_rate": 0.0007890000000000001,
      "loss": 6.3465,
      "step": 263
    },
    {
      "epoch": 0.00264,
      "grad_norm": 0.47601242681958134,
      "learning_rate": 0.0007920000000000001,
      "loss": 6.3294,
      "step": 264
    },
    {
      "epoch": 0.00265,
      "grad_norm": 0.5284160270316348,
      "learning_rate": 0.000795,
      "loss": 6.3216,
      "step": 265
    },
    {
      "epoch": 0.00266,
      "grad_norm": 0.45329071221457967,
      "learning_rate": 0.0007980000000000001,
      "loss": 6.3157,
      "step": 266
    },
    {
      "epoch": 0.00267,
      "grad_norm": 0.5054867442589596,
      "learning_rate": 0.0008010000000000001,
      "loss": 6.3107,
      "step": 267
    },
    {
      "epoch": 0.00268,
      "grad_norm": 0.43889982597834043,
      "learning_rate": 0.000804,
      "loss": 6.3036,
      "step": 268
    },
    {
      "epoch": 0.00269,
      "grad_norm": 0.5126978914131535,
      "learning_rate": 0.0008070000000000001,
      "loss": 6.2878,
      "step": 269
    },
    {
      "epoch": 0.0027,
      "grad_norm": 0.49623237637976697,
      "learning_rate": 0.0008100000000000001,
      "loss": 6.2881,
      "step": 270
    },
    {
      "epoch": 0.00271,
      "grad_norm": 0.44574618013664674,
      "learning_rate": 0.000813,
      "loss": 6.2929,
      "step": 271
    },
    {
      "epoch": 0.00272,
      "grad_norm": 0.5373388276032048,
      "learning_rate": 0.0008160000000000001,
      "loss": 6.2781,
      "step": 272
    },
    {
      "epoch": 0.00273,
      "grad_norm": 0.5925104788121834,
      "learning_rate": 0.0008190000000000001,
      "loss": 6.2673,
      "step": 273
    },
    {
      "epoch": 0.00274,
      "grad_norm": 0.7812486674134582,
      "learning_rate": 0.000822,
      "loss": 6.2537,
      "step": 274
    },
    {
      "epoch": 0.00275,
      "grad_norm": 1.4372466998888491,
      "learning_rate": 0.0008250000000000001,
      "loss": 6.2838,
      "step": 275
    },
    {
      "epoch": 0.00276,
      "grad_norm": 1.009790255288434,
      "learning_rate": 0.0008280000000000001,
      "loss": 6.2706,
      "step": 276
    },
    {
      "epoch": 0.00277,
      "grad_norm": 1.1366790246792218,
      "learning_rate": 0.0008310000000000001,
      "loss": 6.2794,
      "step": 277
    },
    {
      "epoch": 0.00278,
      "grad_norm": 0.7426000684998063,
      "learning_rate": 0.0008340000000000001,
      "loss": 6.2448,
      "step": 278
    },
    {
      "epoch": 0.00279,
      "grad_norm": 1.1507422621364256,
      "learning_rate": 0.0008370000000000001,
      "loss": 6.2577,
      "step": 279
    },
    {
      "epoch": 0.0028,
      "grad_norm": 0.9287326777762811,
      "learning_rate": 0.0008400000000000001,
      "loss": 6.2422,
      "step": 280
    },
    {
      "epoch": 0.00281,
      "grad_norm": 0.6961892841552818,
      "learning_rate": 0.0008430000000000001,
      "loss": 6.2248,
      "step": 281
    },
    {
      "epoch": 0.00282,
      "grad_norm": 0.5479119740633785,
      "learning_rate": 0.000846,
      "loss": 6.219,
      "step": 282
    },
    {
      "epoch": 0.00283,
      "grad_norm": 0.533852757527586,
      "learning_rate": 0.0008489999999999999,
      "loss": 6.2128,
      "step": 283
    },
    {
      "epoch": 0.00284,
      "grad_norm": 0.4914173114643737,
      "learning_rate": 0.0008519999999999999,
      "loss": 6.2012,
      "step": 284
    },
    {
      "epoch": 0.00285,
      "grad_norm": 0.5452017206249536,
      "learning_rate": 0.000855,
      "loss": 6.2041,
      "step": 285
    },
    {
      "epoch": 0.00286,
      "grad_norm": 0.5575344533377293,
      "learning_rate": 0.0008579999999999999,
      "loss": 6.1806,
      "step": 286
    },
    {
      "epoch": 0.00287,
      "grad_norm": 0.5864183738682309,
      "learning_rate": 0.000861,
      "loss": 6.1845,
      "step": 287
    },
    {
      "epoch": 0.00288,
      "grad_norm": 0.5841667976006761,
      "learning_rate": 0.000864,
      "loss": 6.1638,
      "step": 288
    },
    {
      "epoch": 0.00289,
      "grad_norm": 0.46053068944693165,
      "learning_rate": 0.0008669999999999999,
      "loss": 6.1757,
      "step": 289
    },
    {
      "epoch": 0.0029,
      "grad_norm": 0.3129456576237599,
      "learning_rate": 0.00087,
      "loss": 6.1522,
      "step": 290
    },
    {
      "epoch": 0.00291,
      "grad_norm": 0.42419235273038636,
      "learning_rate": 0.000873,
      "loss": 6.1475,
      "step": 291
    },
    {
      "epoch": 0.00292,
      "grad_norm": 0.4195983642240095,
      "learning_rate": 0.0008759999999999999,
      "loss": 6.1444,
      "step": 292
    },
    {
      "epoch": 0.00293,
      "grad_norm": 0.4481246758453938,
      "learning_rate": 0.000879,
      "loss": 6.1262,
      "step": 293
    },
    {
      "epoch": 0.00294,
      "grad_norm": 0.6187999235349624,
      "learning_rate": 0.000882,
      "loss": 6.1427,
      "step": 294
    },
    {
      "epoch": 0.00295,
      "grad_norm": 0.9782573637293742,
      "learning_rate": 0.0008849999999999999,
      "loss": 6.1183,
      "step": 295
    },
    {
      "epoch": 0.00296,
      "grad_norm": 1.1963373758150497,
      "learning_rate": 0.000888,
      "loss": 6.1501,
      "step": 296
    },
    {
      "epoch": 0.00297,
      "grad_norm": 0.599998736767153,
      "learning_rate": 0.000891,
      "loss": 6.0984,
      "step": 297
    },
    {
      "epoch": 0.00298,
      "grad_norm": 0.8013796512698059,
      "learning_rate": 0.0008939999999999999,
      "loss": 6.1246,
      "step": 298
    },
    {
      "epoch": 0.00299,
      "grad_norm": 0.5344179597512946,
      "learning_rate": 0.000897,
      "loss": 6.0868,
      "step": 299
    },
    {
      "epoch": 0.003,
      "grad_norm": 0.7019061294911672,
      "learning_rate": 0.0009,
      "loss": 6.0964,
      "step": 300
    },
    {
      "epoch": 0.00301,
      "grad_norm": 0.6084278233141631,
      "learning_rate": 0.0009029999999999999,
      "loss": 6.0828,
      "step": 301
    },
    {
      "epoch": 0.00302,
      "grad_norm": 0.8187255130373073,
      "learning_rate": 0.000906,
      "loss": 6.0734,
      "step": 302
    },
    {
      "epoch": 0.00303,
      "grad_norm": 0.8857858146685933,
      "learning_rate": 0.000909,
      "loss": 6.0899,
      "step": 303
    },
    {
      "epoch": 0.00304,
      "grad_norm": 1.2136542484144364,
      "learning_rate": 0.000912,
      "loss": 6.0938,
      "step": 304
    },
    {
      "epoch": 0.00305,
      "grad_norm": 1.177396089785494,
      "learning_rate": 0.000915,
      "loss": 6.0972,
      "step": 305
    },
    {
      "epoch": 0.00306,
      "grad_norm": 0.6904788914447576,
      "learning_rate": 0.000918,
      "loss": 6.0696,
      "step": 306
    },
    {
      "epoch": 0.00307,
      "grad_norm": 0.6208733747216669,
      "learning_rate": 0.000921,
      "loss": 6.0635,
      "step": 307
    },
    {
      "epoch": 0.00308,
      "grad_norm": 0.5825923197804476,
      "learning_rate": 0.000924,
      "loss": 6.0448,
      "step": 308
    },
    {
      "epoch": 0.00309,
      "grad_norm": 0.5505228855267935,
      "learning_rate": 0.000927,
      "loss": 6.0605,
      "step": 309
    },
    {
      "epoch": 0.0031,
      "grad_norm": 0.4641714577243504,
      "learning_rate": 0.00093,
      "loss": 6.0377,
      "step": 310
    },
    {
      "epoch": 0.00311,
      "grad_norm": 0.5505721165236824,
      "learning_rate": 0.000933,
      "loss": 6.0314,
      "step": 311
    },
    {
      "epoch": 0.00312,
      "grad_norm": 0.7061898380577587,
      "learning_rate": 0.000936,
      "loss": 6.0144,
      "step": 312
    },
    {
      "epoch": 0.00313,
      "grad_norm": 0.8674813006917282,
      "learning_rate": 0.0009390000000000001,
      "loss": 6.0347,
      "step": 313
    },
    {
      "epoch": 0.00314,
      "grad_norm": 0.9558239920779429,
      "learning_rate": 0.000942,
      "loss": 6.0427,
      "step": 314
    },
    {
      "epoch": 0.00315,
      "grad_norm": 0.8148008359990885,
      "learning_rate": 0.000945,
      "loss": 6.0164,
      "step": 315
    },
    {
      "epoch": 0.00316,
      "grad_norm": 0.8527496297887206,
      "learning_rate": 0.0009480000000000001,
      "loss": 5.992,
      "step": 316
    },
    {
      "epoch": 0.00317,
      "grad_norm": 0.9373289538679535,
      "learning_rate": 0.000951,
      "loss": 6.0135,
      "step": 317
    },
    {
      "epoch": 0.00318,
      "grad_norm": 0.915551727118195,
      "learning_rate": 0.000954,
      "loss": 6.0085,
      "step": 318
    },
    {
      "epoch": 0.00319,
      "grad_norm": 1.0100858429902202,
      "learning_rate": 0.0009570000000000001,
      "loss": 6.0051,
      "step": 319
    },
    {
      "epoch": 0.0032,
      "grad_norm": 0.6230170449860293,
      "learning_rate": 0.00096,
      "loss": 5.9891,
      "step": 320
    },
    {
      "epoch": 0.00321,
      "grad_norm": 0.7031926346795826,
      "learning_rate": 0.000963,
      "loss": 5.9829,
      "step": 321
    },
    {
      "epoch": 0.00322,
      "grad_norm": 0.6443754150070977,
      "learning_rate": 0.0009660000000000001,
      "loss": 5.9655,
      "step": 322
    },
    {
      "epoch": 0.00323,
      "grad_norm": 0.7449741143243878,
      "learning_rate": 0.000969,
      "loss": 5.9659,
      "step": 323
    },
    {
      "epoch": 0.00324,
      "grad_norm": 0.8499498801669653,
      "learning_rate": 0.0009720000000000001,
      "loss": 5.9786,
      "step": 324
    },
    {
      "epoch": 0.00325,
      "grad_norm": 0.7370021274780517,
      "learning_rate": 0.0009750000000000001,
      "loss": 5.9372,
      "step": 325
    },
    {
      "epoch": 0.00326,
      "grad_norm": 0.7692642106759962,
      "learning_rate": 0.0009780000000000001,
      "loss": 5.9561,
      "step": 326
    },
    {
      "epoch": 0.00327,
      "grad_norm": 0.9205755534728525,
      "learning_rate": 0.000981,
      "loss": 5.96,
      "step": 327
    },
    {
      "epoch": 0.00328,
      "grad_norm": 0.9462851248298534,
      "learning_rate": 0.000984,
      "loss": 5.9392,
      "step": 328
    },
    {
      "epoch": 0.00329,
      "grad_norm": 0.6598697228026301,
      "learning_rate": 0.000987,
      "loss": 5.9306,
      "step": 329
    },
    {
      "epoch": 0.0033,
      "grad_norm": 0.6819397096877848,
      "learning_rate": 0.00099,
      "loss": 5.9255,
      "step": 330
    },
    {
      "epoch": 0.00331,
      "grad_norm": 0.7335362675715908,
      "learning_rate": 0.0009930000000000002,
      "loss": 5.9282,
      "step": 331
    },
    {
      "epoch": 0.00332,
      "grad_norm": 0.7128341571534664,
      "learning_rate": 0.0009960000000000001,
      "loss": 5.9115,
      "step": 332
    },
    {
      "epoch": 0.00333,
      "grad_norm": 0.6642964548475624,
      "learning_rate": 0.000999,
      "loss": 5.8958,
      "step": 333
    },
    {
      "epoch": 0.00334,
      "grad_norm": 0.6296302431952895,
      "learning_rate": 0.001002,
      "loss": 5.9043,
      "step": 334
    },
    {
      "epoch": 0.00335,
      "grad_norm": 0.610819189925057,
      "learning_rate": 0.001005,
      "loss": 5.9015,
      "step": 335
    },
    {
      "epoch": 0.00336,
      "grad_norm": 0.5248795967011678,
      "learning_rate": 0.001008,
      "loss": 5.8916,
      "step": 336
    },
    {
      "epoch": 0.00337,
      "grad_norm": 0.474981761201143,
      "learning_rate": 0.0010110000000000002,
      "loss": 5.868,
      "step": 337
    },
    {
      "epoch": 0.00338,
      "grad_norm": 0.4982919042440728,
      "learning_rate": 0.0010140000000000001,
      "loss": 5.8835,
      "step": 338
    },
    {
      "epoch": 0.00339,
      "grad_norm": 0.4565693208938873,
      "learning_rate": 0.0010170000000000001,
      "loss": 5.8849,
      "step": 339
    },
    {
      "epoch": 0.0034,
      "grad_norm": 0.5166741693031021,
      "learning_rate": 0.00102,
      "loss": 5.8551,
      "step": 340
    },
    {
      "epoch": 0.00341,
      "grad_norm": 0.517141824320128,
      "learning_rate": 0.001023,
      "loss": 5.8368,
      "step": 341
    },
    {
      "epoch": 0.00342,
      "grad_norm": 0.5527942444154041,
      "learning_rate": 0.001026,
      "loss": 5.8429,
      "step": 342
    },
    {
      "epoch": 0.00343,
      "grad_norm": 0.6606601975038157,
      "learning_rate": 0.0010290000000000002,
      "loss": 5.8479,
      "step": 343
    },
    {
      "epoch": 0.00344,
      "grad_norm": 0.7784149171541165,
      "learning_rate": 0.001032,
      "loss": 5.8398,
      "step": 344
    },
    {
      "epoch": 0.00345,
      "grad_norm": 0.8397401648868169,
      "learning_rate": 0.001035,
      "loss": 5.8375,
      "step": 345
    },
    {
      "epoch": 0.00346,
      "grad_norm": 0.9944775596372618,
      "learning_rate": 0.0010379999999999999,
      "loss": 5.8449,
      "step": 346
    },
    {
      "epoch": 0.00347,
      "grad_norm": 1.0844541863277397,
      "learning_rate": 0.001041,
      "loss": 5.8271,
      "step": 347
    },
    {
      "epoch": 0.00348,
      "grad_norm": 1.1115702342985518,
      "learning_rate": 0.001044,
      "loss": 5.8384,
      "step": 348
    },
    {
      "epoch": 0.00349,
      "grad_norm": 1.0763373685259212,
      "learning_rate": 0.001047,
      "loss": 5.8194,
      "step": 349
    },
    {
      "epoch": 0.0035,
      "grad_norm": 1.154751185583195,
      "learning_rate": 0.00105,
      "loss": 5.8503,
      "step": 350
    },
    {
      "epoch": 0.00351,
      "grad_norm": 0.9949055141165053,
      "learning_rate": 0.001053,
      "loss": 5.8465,
      "step": 351
    },
    {
      "epoch": 0.00352,
      "grad_norm": 1.1269100337690212,
      "learning_rate": 0.0010559999999999999,
      "loss": 5.8358,
      "step": 352
    },
    {
      "epoch": 0.00353,
      "grad_norm": 0.7896283778005796,
      "learning_rate": 0.001059,
      "loss": 5.8219,
      "step": 353
    },
    {
      "epoch": 0.00354,
      "grad_norm": 0.6875113153469495,
      "learning_rate": 0.001062,
      "loss": 5.7918,
      "step": 354
    },
    {
      "epoch": 0.00355,
      "grad_norm": 0.7321424784274652,
      "learning_rate": 0.001065,
      "loss": 5.8057,
      "step": 355
    },
    {
      "epoch": 0.00356,
      "grad_norm": 0.6838178663323626,
      "learning_rate": 0.001068,
      "loss": 5.7969,
      "step": 356
    },
    {
      "epoch": 0.00357,
      "grad_norm": 0.5941172260432777,
      "learning_rate": 0.001071,
      "loss": 5.7852,
      "step": 357
    },
    {
      "epoch": 0.00358,
      "grad_norm": 0.6211494063140103,
      "learning_rate": 0.001074,
      "loss": 5.7764,
      "step": 358
    },
    {
      "epoch": 0.00359,
      "grad_norm": 0.645459064059718,
      "learning_rate": 0.001077,
      "loss": 5.7674,
      "step": 359
    },
    {
      "epoch": 0.0036,
      "grad_norm": 0.7187316170793214,
      "learning_rate": 0.00108,
      "loss": 5.7713,
      "step": 360
    },
    {
      "epoch": 0.00361,
      "grad_norm": 0.8420575980895092,
      "learning_rate": 0.001083,
      "loss": 5.7694,
      "step": 361
    },
    {
      "epoch": 0.00362,
      "grad_norm": 0.8000647457554158,
      "learning_rate": 0.001086,
      "loss": 5.7526,
      "step": 362
    },
    {
      "epoch": 0.00363,
      "grad_norm": 0.7128868464597033,
      "learning_rate": 0.001089,
      "loss": 5.7444,
      "step": 363
    },
    {
      "epoch": 0.00364,
      "grad_norm": 0.660116265321768,
      "learning_rate": 0.001092,
      "loss": 5.7546,
      "step": 364
    },
    {
      "epoch": 0.00365,
      "grad_norm": 0.7837831658386917,
      "learning_rate": 0.001095,
      "loss": 5.7547,
      "step": 365
    },
    {
      "epoch": 0.00366,
      "grad_norm": 0.7821248144040719,
      "learning_rate": 0.001098,
      "loss": 5.7348,
      "step": 366
    },
    {
      "epoch": 0.00367,
      "grad_norm": 0.8460018101445828,
      "learning_rate": 0.001101,
      "loss": 5.7355,
      "step": 367
    },
    {
      "epoch": 0.00368,
      "grad_norm": 0.962056228227274,
      "learning_rate": 0.001104,
      "loss": 5.7428,
      "step": 368
    },
    {
      "epoch": 0.00369,
      "grad_norm": 0.8851788755380371,
      "learning_rate": 0.001107,
      "loss": 5.7443,
      "step": 369
    },
    {
      "epoch": 0.0037,
      "grad_norm": 0.7906813569461731,
      "learning_rate": 0.00111,
      "loss": 5.7238,
      "step": 370
    },
    {
      "epoch": 0.00371,
      "grad_norm": 0.6753760943719406,
      "learning_rate": 0.001113,
      "loss": 5.6912,
      "step": 371
    },
    {
      "epoch": 0.00372,
      "grad_norm": 0.6449512979931109,
      "learning_rate": 0.001116,
      "loss": 5.712,
      "step": 372
    },
    {
      "epoch": 0.00373,
      "grad_norm": 0.7672882935081258,
      "learning_rate": 0.001119,
      "loss": 5.7038,
      "step": 373
    },
    {
      "epoch": 0.00374,
      "grad_norm": 0.9442109709312425,
      "learning_rate": 0.001122,
      "loss": 5.7133,
      "step": 374
    },
    {
      "epoch": 0.00375,
      "grad_norm": 0.8527878740214185,
      "learning_rate": 0.0011250000000000001,
      "loss": 5.7144,
      "step": 375
    },
    {
      "epoch": 0.00376,
      "grad_norm": 0.6469115537095235,
      "learning_rate": 0.001128,
      "loss": 5.7041,
      "step": 376
    },
    {
      "epoch": 0.00377,
      "grad_norm": 0.6476498140164245,
      "learning_rate": 0.001131,
      "loss": 5.6975,
      "step": 377
    },
    {
      "epoch": 0.00378,
      "grad_norm": 0.6293494723973451,
      "learning_rate": 0.001134,
      "loss": 5.6937,
      "step": 378
    },
    {
      "epoch": 0.00379,
      "grad_norm": 0.5839693486494109,
      "learning_rate": 0.001137,
      "loss": 5.6765,
      "step": 379
    },
    {
      "epoch": 0.0038,
      "grad_norm": 0.6735782782443015,
      "learning_rate": 0.00114,
      "loss": 5.6974,
      "step": 380
    },
    {
      "epoch": 0.00381,
      "grad_norm": 0.8231233374417803,
      "learning_rate": 0.0011430000000000001,
      "loss": 5.6678,
      "step": 381
    },
    {
      "epoch": 0.00382,
      "grad_norm": 0.9310146874963119,
      "learning_rate": 0.001146,
      "loss": 5.6605,
      "step": 382
    },
    {
      "epoch": 0.00383,
      "grad_norm": 0.8465306298864036,
      "learning_rate": 0.001149,
      "loss": 5.6696,
      "step": 383
    },
    {
      "epoch": 0.00384,
      "grad_norm": 0.7735682894418534,
      "learning_rate": 0.001152,
      "loss": 5.6728,
      "step": 384
    },
    {
      "epoch": 0.00385,
      "grad_norm": 0.7108251568334872,
      "learning_rate": 0.001155,
      "loss": 5.66,
      "step": 385
    },
    {
      "epoch": 0.00386,
      "grad_norm": 0.6607218758999341,
      "learning_rate": 0.001158,
      "loss": 5.6625,
      "step": 386
    },
    {
      "epoch": 0.00387,
      "grad_norm": 0.8710168775036707,
      "learning_rate": 0.0011610000000000001,
      "loss": 5.6396,
      "step": 387
    },
    {
      "epoch": 0.00388,
      "grad_norm": 1.2329079335292137,
      "learning_rate": 0.001164,
      "loss": 5.6555,
      "step": 388
    },
    {
      "epoch": 0.00389,
      "grad_norm": 0.9663820489145136,
      "learning_rate": 0.001167,
      "loss": 5.6567,
      "step": 389
    },
    {
      "epoch": 0.0039,
      "grad_norm": 1.1217202209674602,
      "learning_rate": 0.00117,
      "loss": 5.644,
      "step": 390
    },
    {
      "epoch": 0.00391,
      "grad_norm": 1.151101954947749,
      "learning_rate": 0.001173,
      "loss": 5.6576,
      "step": 391
    },
    {
      "epoch": 0.00392,
      "grad_norm": 0.8694084215496658,
      "learning_rate": 0.001176,
      "loss": 5.66,
      "step": 392
    },
    {
      "epoch": 0.00393,
      "grad_norm": 0.9543644853323229,
      "learning_rate": 0.0011790000000000001,
      "loss": 5.6338,
      "step": 393
    },
    {
      "epoch": 0.00394,
      "grad_norm": 0.8780145887874276,
      "learning_rate": 0.001182,
      "loss": 5.6222,
      "step": 394
    },
    {
      "epoch": 0.00395,
      "grad_norm": 0.7527231415908622,
      "learning_rate": 0.001185,
      "loss": 5.6018,
      "step": 395
    },
    {
      "epoch": 0.00396,
      "grad_norm": 0.8774478125599554,
      "learning_rate": 0.001188,
      "loss": 5.6219,
      "step": 396
    },
    {
      "epoch": 0.00397,
      "grad_norm": 0.9175690164934137,
      "learning_rate": 0.001191,
      "loss": 5.6177,
      "step": 397
    },
    {
      "epoch": 0.00398,
      "grad_norm": 0.7008521842282015,
      "learning_rate": 0.0011940000000000002,
      "loss": 5.6167,
      "step": 398
    },
    {
      "epoch": 0.00399,
      "grad_norm": 0.7914810239489021,
      "learning_rate": 0.0011970000000000001,
      "loss": 5.6198,
      "step": 399
    },
    {
      "epoch": 0.004,
      "grad_norm": 0.8769542296732423,
      "learning_rate": 0.0012000000000000001,
      "loss": 5.6078,
      "step": 400
    },
    {
      "epoch": 0.00401,
      "grad_norm": 0.9532232290382274,
      "learning_rate": 0.001203,
      "loss": 5.5929,
      "step": 401
    },
    {
      "epoch": 0.00402,
      "grad_norm": 0.7358311359997632,
      "learning_rate": 0.001206,
      "loss": 5.5932,
      "step": 402
    },
    {
      "epoch": 0.00403,
      "grad_norm": 0.5843625246076212,
      "learning_rate": 0.001209,
      "loss": 5.5883,
      "step": 403
    },
    {
      "epoch": 0.00404,
      "grad_norm": 0.5438164721501143,
      "learning_rate": 0.0012120000000000002,
      "loss": 5.5786,
      "step": 404
    },
    {
      "epoch": 0.00405,
      "grad_norm": 0.503935889546328,
      "learning_rate": 0.0012150000000000002,
      "loss": 5.5729,
      "step": 405
    },
    {
      "epoch": 0.00406,
      "grad_norm": 0.581555817136203,
      "learning_rate": 0.0012180000000000001,
      "loss": 5.5609,
      "step": 406
    },
    {
      "epoch": 0.00407,
      "grad_norm": 0.6908000438827393,
      "learning_rate": 0.0012209999999999999,
      "loss": 5.5552,
      "step": 407
    },
    {
      "epoch": 0.00408,
      "grad_norm": 0.9017213905273226,
      "learning_rate": 0.001224,
      "loss": 5.5815,
      "step": 408
    },
    {
      "epoch": 0.00409,
      "grad_norm": 0.8946003332710046,
      "learning_rate": 0.001227,
      "loss": 5.5595,
      "step": 409
    },
    {
      "epoch": 0.0041,
      "grad_norm": 0.8697402094229448,
      "learning_rate": 0.00123,
      "loss": 5.5572,
      "step": 410
    },
    {
      "epoch": 0.00411,
      "grad_norm": 0.889615246973329,
      "learning_rate": 0.001233,
      "loss": 5.5739,
      "step": 411
    },
    {
      "epoch": 0.00412,
      "grad_norm": 1.0125641304675919,
      "learning_rate": 0.001236,
      "loss": 5.5666,
      "step": 412
    },
    {
      "epoch": 0.00413,
      "grad_norm": 0.8779023797874608,
      "learning_rate": 0.0012389999999999999,
      "loss": 5.5643,
      "step": 413
    },
    {
      "epoch": 0.00414,
      "grad_norm": 0.5508909189800907,
      "learning_rate": 0.001242,
      "loss": 5.5325,
      "step": 414
    },
    {
      "epoch": 0.00415,
      "grad_norm": 0.5649392842818297,
      "learning_rate": 0.001245,
      "loss": 5.5358,
      "step": 415
    },
    {
      "epoch": 0.00416,
      "grad_norm": 0.5511318314324237,
      "learning_rate": 0.001248,
      "loss": 5.5166,
      "step": 416
    },
    {
      "epoch": 0.00417,
      "grad_norm": 0.5999723506722826,
      "learning_rate": 0.001251,
      "loss": 5.5477,
      "step": 417
    },
    {
      "epoch": 0.00418,
      "grad_norm": 0.6470183959659118,
      "learning_rate": 0.001254,
      "loss": 5.5224,
      "step": 418
    },
    {
      "epoch": 0.00419,
      "grad_norm": 0.7320078508628892,
      "learning_rate": 0.0012569999999999999,
      "loss": 5.5107,
      "step": 419
    },
    {
      "epoch": 0.0042,
      "grad_norm": 0.8456080065384579,
      "learning_rate": 0.00126,
      "loss": 5.5293,
      "step": 420
    },
    {
      "epoch": 0.00421,
      "grad_norm": 0.9939084210290385,
      "learning_rate": 0.001263,
      "loss": 5.5281,
      "step": 421
    },
    {
      "epoch": 0.00422,
      "grad_norm": 1.1479664785212111,
      "learning_rate": 0.001266,
      "loss": 5.5264,
      "step": 422
    },
    {
      "epoch": 0.00423,
      "grad_norm": 1.0903698179641128,
      "learning_rate": 0.001269,
      "loss": 5.5217,
      "step": 423
    },
    {
      "epoch": 0.00424,
      "grad_norm": 1.0214513010937782,
      "learning_rate": 0.001272,
      "loss": 5.5248,
      "step": 424
    },
    {
      "epoch": 0.00425,
      "grad_norm": 1.034943759284661,
      "learning_rate": 0.001275,
      "loss": 5.5311,
      "step": 425
    },
    {
      "epoch": 0.00426,
      "grad_norm": 0.7614861129943166,
      "learning_rate": 0.001278,
      "loss": 5.5212,
      "step": 426
    },
    {
      "epoch": 0.00427,
      "grad_norm": 0.7672748957608901,
      "learning_rate": 0.001281,
      "loss": 5.5058,
      "step": 427
    },
    {
      "epoch": 0.00428,
      "grad_norm": 0.8987012131115926,
      "learning_rate": 0.001284,
      "loss": 5.4868,
      "step": 428
    },
    {
      "epoch": 0.00429,
      "grad_norm": 0.8700946902461411,
      "learning_rate": 0.001287,
      "loss": 5.5045,
      "step": 429
    },
    {
      "epoch": 0.0043,
      "grad_norm": 0.9264807594834162,
      "learning_rate": 0.00129,
      "loss": 5.5003,
      "step": 430
    },
    {
      "epoch": 0.00431,
      "grad_norm": 0.9714177447726252,
      "learning_rate": 0.001293,
      "loss": 5.4888,
      "step": 431
    },
    {
      "epoch": 0.00432,
      "grad_norm": 0.8362399273583541,
      "learning_rate": 0.001296,
      "loss": 5.498,
      "step": 432
    },
    {
      "epoch": 0.00433,
      "grad_norm": 0.7360288767544512,
      "learning_rate": 0.001299,
      "loss": 5.4879,
      "step": 433
    },
    {
      "epoch": 0.00434,
      "grad_norm": 0.49506730909723906,
      "learning_rate": 0.001302,
      "loss": 5.4769,
      "step": 434
    },
    {
      "epoch": 0.00435,
      "grad_norm": 0.5541770965077963,
      "learning_rate": 0.001305,
      "loss": 5.4763,
      "step": 435
    },
    {
      "epoch": 0.00436,
      "grad_norm": 0.4143567360382353,
      "learning_rate": 0.001308,
      "loss": 5.448,
      "step": 436
    },
    {
      "epoch": 0.00437,
      "grad_norm": 0.5010974412085327,
      "learning_rate": 0.001311,
      "loss": 5.468,
      "step": 437
    },
    {
      "epoch": 0.00438,
      "grad_norm": 0.4773129501944819,
      "learning_rate": 0.001314,
      "loss": 5.4471,
      "step": 438
    },
    {
      "epoch": 0.00439,
      "grad_norm": 0.4480238817975887,
      "learning_rate": 0.001317,
      "loss": 5.4733,
      "step": 439
    },
    {
      "epoch": 0.0044,
      "grad_norm": 0.5102675481146428,
      "learning_rate": 0.00132,
      "loss": 5.4437,
      "step": 440
    },
    {
      "epoch": 0.00441,
      "grad_norm": 0.7603199484002093,
      "learning_rate": 0.001323,
      "loss": 5.4453,
      "step": 441
    },
    {
      "epoch": 0.00442,
      "grad_norm": 1.1458651313729942,
      "learning_rate": 0.0013260000000000001,
      "loss": 5.4537,
      "step": 442
    },
    {
      "epoch": 0.00443,
      "grad_norm": 0.8849689457299824,
      "learning_rate": 0.001329,
      "loss": 5.4514,
      "step": 443
    },
    {
      "epoch": 0.00444,
      "grad_norm": 0.8143276002862829,
      "learning_rate": 0.001332,
      "loss": 5.4477,
      "step": 444
    },
    {
      "epoch": 0.00445,
      "grad_norm": 1.014187478097306,
      "learning_rate": 0.001335,
      "loss": 5.4432,
      "step": 445
    },
    {
      "epoch": 0.00446,
      "grad_norm": 1.0652502023957744,
      "learning_rate": 0.001338,
      "loss": 5.449,
      "step": 446
    },
    {
      "epoch": 0.00447,
      "grad_norm": 0.9667242613798618,
      "learning_rate": 0.001341,
      "loss": 5.4482,
      "step": 447
    },
    {
      "epoch": 0.00448,
      "grad_norm": 0.8366191486526497,
      "learning_rate": 0.0013440000000000001,
      "loss": 5.4279,
      "step": 448
    },
    {
      "epoch": 0.00449,
      "grad_norm": 0.7799129698781961,
      "learning_rate": 0.001347,
      "loss": 5.43,
      "step": 449
    },
    {
      "epoch": 0.0045,
      "grad_norm": 0.7012481729938261,
      "learning_rate": 0.00135,
      "loss": 5.4286,
      "step": 450
    },
    {
      "epoch": 0.00451,
      "grad_norm": 0.7248552907075281,
      "learning_rate": 0.001353,
      "loss": 5.4198,
      "step": 451
    },
    {
      "epoch": 0.00452,
      "grad_norm": 0.7044649784518014,
      "learning_rate": 0.001356,
      "loss": 5.4008,
      "step": 452
    },
    {
      "epoch": 0.00453,
      "grad_norm": 0.7484508328215168,
      "learning_rate": 0.001359,
      "loss": 5.3954,
      "step": 453
    },
    {
      "epoch": 0.00454,
      "grad_norm": 0.8395085041820342,
      "learning_rate": 0.0013620000000000001,
      "loss": 5.4145,
      "step": 454
    },
    {
      "epoch": 0.00455,
      "grad_norm": 0.8415231105291262,
      "learning_rate": 0.0013650000000000001,
      "loss": 5.4127,
      "step": 455
    },
    {
      "epoch": 0.00456,
      "grad_norm": 0.7255220872264407,
      "learning_rate": 0.001368,
      "loss": 5.4058,
      "step": 456
    },
    {
      "epoch": 0.00457,
      "grad_norm": 0.9594202170388311,
      "learning_rate": 0.001371,
      "loss": 5.4022,
      "step": 457
    },
    {
      "epoch": 0.00458,
      "grad_norm": 1.3293728411767247,
      "learning_rate": 0.001374,
      "loss": 5.432,
      "step": 458
    },
    {
      "epoch": 0.00459,
      "grad_norm": 0.6757510453727773,
      "learning_rate": 0.0013770000000000002,
      "loss": 5.3986,
      "step": 459
    },
    {
      "epoch": 0.0046,
      "grad_norm": 0.642286715349539,
      "learning_rate": 0.0013800000000000002,
      "loss": 5.3842,
      "step": 460
    },
    {
      "epoch": 0.00461,
      "grad_norm": 0.718506682267116,
      "learning_rate": 0.0013830000000000001,
      "loss": 5.3809,
      "step": 461
    },
    {
      "epoch": 0.00462,
      "grad_norm": 0.8721301960351167,
      "learning_rate": 0.001386,
      "loss": 5.3962,
      "step": 462
    },
    {
      "epoch": 0.00463,
      "grad_norm": 1.0129389296679803,
      "learning_rate": 0.001389,
      "loss": 5.378,
      "step": 463
    },
    {
      "epoch": 0.00464,
      "grad_norm": 1.0088046974816025,
      "learning_rate": 0.001392,
      "loss": 5.4001,
      "step": 464
    },
    {
      "epoch": 0.00465,
      "grad_norm": 1.0306323020829953,
      "learning_rate": 0.0013950000000000002,
      "loss": 5.3734,
      "step": 465
    },
    {
      "epoch": 0.00466,
      "grad_norm": 0.8720845065856513,
      "learning_rate": 0.0013980000000000002,
      "loss": 5.3874,
      "step": 466
    },
    {
      "epoch": 0.00467,
      "grad_norm": 0.8887797743014489,
      "learning_rate": 0.0014010000000000001,
      "loss": 5.3768,
      "step": 467
    },
    {
      "epoch": 0.00468,
      "grad_norm": 0.9259005236029753,
      "learning_rate": 0.001404,
      "loss": 5.3729,
      "step": 468
    },
    {
      "epoch": 0.00469,
      "grad_norm": 0.8792163356951412,
      "learning_rate": 0.001407,
      "loss": 5.3748,
      "step": 469
    },
    {
      "epoch": 0.0047,
      "grad_norm": 0.9535575619633974,
      "learning_rate": 0.00141,
      "loss": 5.3888,
      "step": 470
    },
    {
      "epoch": 0.00471,
      "grad_norm": 0.8943151432047335,
      "learning_rate": 0.001413,
      "loss": 5.3656,
      "step": 471
    },
    {
      "epoch": 0.00472,
      "grad_norm": 0.9438166210457617,
      "learning_rate": 0.001416,
      "loss": 5.3818,
      "step": 472
    },
    {
      "epoch": 0.00473,
      "grad_norm": 0.955062498554473,
      "learning_rate": 0.001419,
      "loss": 5.3925,
      "step": 473
    },
    {
      "epoch": 0.00474,
      "grad_norm": 0.8744713005253077,
      "learning_rate": 0.0014219999999999999,
      "loss": 5.3698,
      "step": 474
    },
    {
      "epoch": 0.00475,
      "grad_norm": 0.9433538313620896,
      "learning_rate": 0.001425,
      "loss": 5.3673,
      "step": 475
    },
    {
      "epoch": 0.00476,
      "grad_norm": 1.003831968151295,
      "learning_rate": 0.001428,
      "loss": 5.4007,
      "step": 476
    },
    {
      "epoch": 0.00477,
      "grad_norm": 1.1933630822112062,
      "learning_rate": 0.001431,
      "loss": 5.3632,
      "step": 477
    },
    {
      "epoch": 0.00478,
      "grad_norm": 0.862117250818067,
      "learning_rate": 0.001434,
      "loss": 5.3781,
      "step": 478
    },
    {
      "epoch": 0.00479,
      "grad_norm": 0.8127073460113728,
      "learning_rate": 0.001437,
      "loss": 5.3546,
      "step": 479
    },
    {
      "epoch": 0.0048,
      "grad_norm": 0.7646068732813385,
      "learning_rate": 0.0014399999999999999,
      "loss": 5.3463,
      "step": 480
    },
    {
      "epoch": 0.00481,
      "grad_norm": 0.7199260331938973,
      "learning_rate": 0.001443,
      "loss": 5.3466,
      "step": 481
    },
    {
      "epoch": 0.00482,
      "grad_norm": 0.6802449871429381,
      "learning_rate": 0.001446,
      "loss": 5.3202,
      "step": 482
    },
    {
      "epoch": 0.00483,
      "grad_norm": 0.6495861656689597,
      "learning_rate": 0.001449,
      "loss": 5.3291,
      "step": 483
    },
    {
      "epoch": 0.00484,
      "grad_norm": 0.644169893492728,
      "learning_rate": 0.001452,
      "loss": 5.3277,
      "step": 484
    },
    {
      "epoch": 0.00485,
      "grad_norm": 0.7933051477323634,
      "learning_rate": 0.001455,
      "loss": 5.3368,
      "step": 485
    },
    {
      "epoch": 0.00486,
      "grad_norm": 0.8700253631116802,
      "learning_rate": 0.001458,
      "loss": 5.3096,
      "step": 486
    },
    {
      "epoch": 0.00487,
      "grad_norm": 0.7948159001624678,
      "learning_rate": 0.001461,
      "loss": 5.3257,
      "step": 487
    },
    {
      "epoch": 0.00488,
      "grad_norm": 0.5603690187831122,
      "learning_rate": 0.001464,
      "loss": 5.294,
      "step": 488
    },
    {
      "epoch": 0.00489,
      "grad_norm": 0.6185202940839952,
      "learning_rate": 0.001467,
      "loss": 5.287,
      "step": 489
    },
    {
      "epoch": 0.0049,
      "grad_norm": 0.5616341711444851,
      "learning_rate": 0.00147,
      "loss": 5.2988,
      "step": 490
    },
    {
      "epoch": 0.00491,
      "grad_norm": 0.49889379281960217,
      "learning_rate": 0.001473,
      "loss": 5.2927,
      "step": 491
    },
    {
      "epoch": 0.00492,
      "grad_norm": 0.5492592192725451,
      "learning_rate": 0.001476,
      "loss": 5.2915,
      "step": 492
    },
    {
      "epoch": 0.00493,
      "grad_norm": 0.5409556054523526,
      "learning_rate": 0.001479,
      "loss": 5.2842,
      "step": 493
    },
    {
      "epoch": 0.00494,
      "grad_norm": 0.5986439471872183,
      "learning_rate": 0.001482,
      "loss": 5.29,
      "step": 494
    },
    {
      "epoch": 0.00495,
      "grad_norm": 0.5846070904075265,
      "learning_rate": 0.001485,
      "loss": 5.2961,
      "step": 495
    },
    {
      "epoch": 0.00496,
      "grad_norm": 0.5348806476926008,
      "learning_rate": 0.001488,
      "loss": 5.2784,
      "step": 496
    },
    {
      "epoch": 0.00497,
      "grad_norm": 0.4761317629398082,
      "learning_rate": 0.001491,
      "loss": 5.2561,
      "step": 497
    },
    {
      "epoch": 0.00498,
      "grad_norm": 0.5515846127270029,
      "learning_rate": 0.001494,
      "loss": 5.2847,
      "step": 498
    },
    {
      "epoch": 0.00499,
      "grad_norm": 0.5172721915151642,
      "learning_rate": 0.001497,
      "loss": 5.2576,
      "step": 499
    },
    {
      "epoch": 0.005,
      "grad_norm": 0.5199821616229346,
      "learning_rate": 0.0015,
      "loss": 5.2619,
      "step": 500
    },
    {
      "epoch": 0.00501,
      "grad_norm": 0.6001320721934307,
      "learning_rate": 0.001503,
      "loss": 5.2644,
      "step": 501
    },
    {
      "epoch": 0.00502,
      "grad_norm": 0.6030864866826301,
      "learning_rate": 0.001506,
      "loss": 5.2703,
      "step": 502
    },
    {
      "epoch": 0.00503,
      "grad_norm": 0.5191760166109092,
      "learning_rate": 0.0015090000000000001,
      "loss": 5.2717,
      "step": 503
    },
    {
      "epoch": 0.00504,
      "grad_norm": 0.5783228415677168,
      "learning_rate": 0.001512,
      "loss": 5.254,
      "step": 504
    },
    {
      "epoch": 0.00505,
      "grad_norm": 0.6689655395334532,
      "learning_rate": 0.001515,
      "loss": 5.2492,
      "step": 505
    },
    {
      "epoch": 0.00506,
      "grad_norm": 0.6647691433281518,
      "learning_rate": 0.001518,
      "loss": 5.2491,
      "step": 506
    },
    {
      "epoch": 0.00507,
      "grad_norm": 0.8889512182842091,
      "learning_rate": 0.001521,
      "loss": 5.2524,
      "step": 507
    },
    {
      "epoch": 0.00508,
      "grad_norm": 0.890408095664645,
      "learning_rate": 0.001524,
      "loss": 5.258,
      "step": 508
    },
    {
      "epoch": 0.00509,
      "grad_norm": 0.7970662574337232,
      "learning_rate": 0.0015270000000000001,
      "loss": 5.2359,
      "step": 509
    },
    {
      "epoch": 0.0051,
      "grad_norm": 1.2307905246856345,
      "learning_rate": 0.0015300000000000001,
      "loss": 5.2751,
      "step": 510
    },
    {
      "epoch": 0.00511,
      "grad_norm": 0.7669806359491947,
      "learning_rate": 0.001533,
      "loss": 5.2366,
      "step": 511
    },
    {
      "epoch": 0.00512,
      "grad_norm": 0.791675505211477,
      "learning_rate": 0.001536,
      "loss": 5.2387,
      "step": 512
    },
    {
      "epoch": 0.00513,
      "grad_norm": 0.8098770401540116,
      "learning_rate": 0.001539,
      "loss": 5.2533,
      "step": 513
    },
    {
      "epoch": 0.00514,
      "grad_norm": 0.9791401462579269,
      "learning_rate": 0.001542,
      "loss": 5.2434,
      "step": 514
    },
    {
      "epoch": 0.00515,
      "grad_norm": 1.2171542206910853,
      "learning_rate": 0.0015450000000000001,
      "loss": 5.2622,
      "step": 515
    },
    {
      "epoch": 0.00516,
      "grad_norm": 1.023390089785043,
      "learning_rate": 0.0015480000000000001,
      "loss": 5.2555,
      "step": 516
    },
    {
      "epoch": 0.00517,
      "grad_norm": 1.1466434595757542,
      "learning_rate": 0.001551,
      "loss": 5.2575,
      "step": 517
    },
    {
      "epoch": 0.00518,
      "grad_norm": 0.8131985942906488,
      "learning_rate": 0.001554,
      "loss": 5.2332,
      "step": 518
    },
    {
      "epoch": 0.00519,
      "grad_norm": 1.0287267638756976,
      "learning_rate": 0.001557,
      "loss": 5.2441,
      "step": 519
    },
    {
      "epoch": 0.0052,
      "grad_norm": 1.2674528356226307,
      "learning_rate": 0.0015600000000000002,
      "loss": 5.2693,
      "step": 520
    },
    {
      "epoch": 0.00521,
      "grad_norm": 1.0294461133705461,
      "learning_rate": 0.0015630000000000002,
      "loss": 5.2367,
      "step": 521
    },
    {
      "epoch": 0.00522,
      "grad_norm": 0.9591121151827766,
      "learning_rate": 0.0015660000000000001,
      "loss": 5.2358,
      "step": 522
    },
    {
      "epoch": 0.00523,
      "grad_norm": 1.1408114706321955,
      "learning_rate": 0.001569,
      "loss": 5.2593,
      "step": 523
    },
    {
      "epoch": 0.00524,
      "grad_norm": 1.0111329673862832,
      "learning_rate": 0.001572,
      "loss": 5.2618,
      "step": 524
    },
    {
      "epoch": 0.00525,
      "grad_norm": 0.9661016074260314,
      "learning_rate": 0.001575,
      "loss": 5.2564,
      "step": 525
    },
    {
      "epoch": 0.00526,
      "grad_norm": 0.843793078461792,
      "learning_rate": 0.0015780000000000002,
      "loss": 5.2213,
      "step": 526
    },
    {
      "epoch": 0.00527,
      "grad_norm": 0.9400473384736733,
      "learning_rate": 0.0015810000000000002,
      "loss": 5.2433,
      "step": 527
    },
    {
      "epoch": 0.00528,
      "grad_norm": 1.243261864721202,
      "learning_rate": 0.0015840000000000001,
      "loss": 5.2328,
      "step": 528
    },
    {
      "epoch": 0.00529,
      "grad_norm": 0.8157117130861912,
      "learning_rate": 0.001587,
      "loss": 5.2295,
      "step": 529
    },
    {
      "epoch": 0.0053,
      "grad_norm": 0.7227236931041067,
      "learning_rate": 0.00159,
      "loss": 5.2376,
      "step": 530
    },
    {
      "epoch": 0.00531,
      "grad_norm": 0.7295638878978784,
      "learning_rate": 0.001593,
      "loss": 5.2183,
      "step": 531
    },
    {
      "epoch": 0.00532,
      "grad_norm": 1.0309756013631586,
      "learning_rate": 0.0015960000000000002,
      "loss": 5.2317,
      "step": 532
    },
    {
      "epoch": 0.00533,
      "grad_norm": 1.341295235532221,
      "learning_rate": 0.0015990000000000002,
      "loss": 5.2489,
      "step": 533
    },
    {
      "epoch": 0.00534,
      "grad_norm": 0.5418763696970031,
      "learning_rate": 0.0016020000000000001,
      "loss": 5.2102,
      "step": 534
    },
    {
      "epoch": 0.00535,
      "grad_norm": 0.6280107905402296,
      "learning_rate": 0.001605,
      "loss": 5.2186,
      "step": 535
    },
    {
      "epoch": 0.00536,
      "grad_norm": 0.6018626261301364,
      "learning_rate": 0.001608,
      "loss": 5.2146,
      "step": 536
    },
    {
      "epoch": 0.00537,
      "grad_norm": 0.7029103701631062,
      "learning_rate": 0.0016110000000000002,
      "loss": 5.2095,
      "step": 537
    },
    {
      "epoch": 0.00538,
      "grad_norm": 0.6704768823630471,
      "learning_rate": 0.0016140000000000002,
      "loss": 5.1871,
      "step": 538
    },
    {
      "epoch": 0.00539,
      "grad_norm": 0.5297125620314311,
      "learning_rate": 0.0016170000000000002,
      "loss": 5.1804,
      "step": 539
    },
    {
      "epoch": 0.0054,
      "grad_norm": 0.5293869004936285,
      "learning_rate": 0.0016200000000000001,
      "loss": 5.1861,
      "step": 540
    },
    {
      "epoch": 0.00541,
      "grad_norm": 0.6423736030714277,
      "learning_rate": 0.001623,
      "loss": 5.1794,
      "step": 541
    },
    {
      "epoch": 0.00542,
      "grad_norm": 0.7347991006328795,
      "learning_rate": 0.001626,
      "loss": 5.2023,
      "step": 542
    },
    {
      "epoch": 0.00543,
      "grad_norm": 0.8573526470960469,
      "learning_rate": 0.0016290000000000002,
      "loss": 5.1766,
      "step": 543
    },
    {
      "epoch": 0.00544,
      "grad_norm": 0.8407261490154571,
      "learning_rate": 0.0016320000000000002,
      "loss": 5.1927,
      "step": 544
    },
    {
      "epoch": 0.00545,
      "grad_norm": 0.880652021609057,
      "learning_rate": 0.0016350000000000002,
      "loss": 5.2041,
      "step": 545
    },
    {
      "epoch": 0.00546,
      "grad_norm": 0.8052038715443169,
      "learning_rate": 0.0016380000000000001,
      "loss": 5.1733,
      "step": 546
    },
    {
      "epoch": 0.00547,
      "grad_norm": 1.039008513503308,
      "learning_rate": 0.001641,
      "loss": 5.1801,
      "step": 547
    },
    {
      "epoch": 0.00548,
      "grad_norm": 0.9852213695140999,
      "learning_rate": 0.001644,
      "loss": 5.1858,
      "step": 548
    },
    {
      "epoch": 0.00549,
      "grad_norm": 0.8484168198653431,
      "learning_rate": 0.0016470000000000002,
      "loss": 5.1761,
      "step": 549
    },
    {
      "epoch": 0.0055,
      "grad_norm": 0.7806834244324883,
      "learning_rate": 0.0016500000000000002,
      "loss": 5.1846,
      "step": 550
    },
    {
      "epoch": 0.00551,
      "grad_norm": 0.84417611786463,
      "learning_rate": 0.0016530000000000002,
      "loss": 5.1857,
      "step": 551
    },
    {
      "epoch": 0.00552,
      "grad_norm": 0.7190100339826029,
      "learning_rate": 0.0016560000000000001,
      "loss": 5.1737,
      "step": 552
    },
    {
      "epoch": 0.00553,
      "grad_norm": 0.8870015639864556,
      "learning_rate": 0.001659,
      "loss": 5.1694,
      "step": 553
    },
    {
      "epoch": 0.00554,
      "grad_norm": 0.9374486447018174,
      "learning_rate": 0.0016620000000000003,
      "loss": 5.1739,
      "step": 554
    },
    {
      "epoch": 0.00555,
      "grad_norm": 0.8845970130105012,
      "learning_rate": 0.0016650000000000002,
      "loss": 5.1752,
      "step": 555
    },
    {
      "epoch": 0.00556,
      "grad_norm": 1.1195859643592967,
      "learning_rate": 0.0016680000000000002,
      "loss": 5.175,
      "step": 556
    },
    {
      "epoch": 0.00557,
      "grad_norm": 0.7291605041068904,
      "learning_rate": 0.0016710000000000002,
      "loss": 5.1436,
      "step": 557
    },
    {
      "epoch": 0.00558,
      "grad_norm": 0.5874484008921741,
      "learning_rate": 0.0016740000000000001,
      "loss": 5.1692,
      "step": 558
    },
    {
      "epoch": 0.00559,
      "grad_norm": 0.5085539730956898,
      "learning_rate": 0.001677,
      "loss": 5.1457,
      "step": 559
    },
    {
      "epoch": 0.0056,
      "grad_norm": 0.5291508832604556,
      "learning_rate": 0.0016800000000000003,
      "loss": 5.1235,
      "step": 560
    },
    {
      "epoch": 0.00561,
      "grad_norm": 0.5189974837553842,
      "learning_rate": 0.0016830000000000003,
      "loss": 5.1467,
      "step": 561
    },
    {
      "epoch": 0.00562,
      "grad_norm": 0.4777820071853692,
      "learning_rate": 0.0016860000000000002,
      "loss": 5.1356,
      "step": 562
    },
    {
      "epoch": 0.00563,
      "grad_norm": 0.504889791549622,
      "learning_rate": 0.001689,
      "loss": 5.1357,
      "step": 563
    },
    {
      "epoch": 0.00564,
      "grad_norm": 0.5778473577877967,
      "learning_rate": 0.001692,
      "loss": 5.1198,
      "step": 564
    },
    {
      "epoch": 0.00565,
      "grad_norm": 0.7633412851296998,
      "learning_rate": 0.001695,
      "loss": 5.1252,
      "step": 565
    },
    {
      "epoch": 0.00566,
      "grad_norm": 0.8414122298345739,
      "learning_rate": 0.0016979999999999999,
      "loss": 5.1285,
      "step": 566
    },
    {
      "epoch": 0.00567,
      "grad_norm": 0.6675623145494364,
      "learning_rate": 0.0017009999999999998,
      "loss": 5.1143,
      "step": 567
    },
    {
      "epoch": 0.00568,
      "grad_norm": 0.7239302836610552,
      "learning_rate": 0.0017039999999999998,
      "loss": 5.1279,
      "step": 568
    },
    {
      "epoch": 0.00569,
      "grad_norm": 0.811387127406939,
      "learning_rate": 0.001707,
      "loss": 5.1137,
      "step": 569
    },
    {
      "epoch": 0.0057,
      "grad_norm": 0.7754716722905325,
      "learning_rate": 0.00171,
      "loss": 5.1282,
      "step": 570
    },
    {
      "epoch": 0.00571,
      "grad_norm": 0.8562307066263974,
      "learning_rate": 0.001713,
      "loss": 5.1309,
      "step": 571
    },
    {
      "epoch": 0.00572,
      "grad_norm": 1.0627285671250943,
      "learning_rate": 0.0017159999999999999,
      "loss": 5.1191,
      "step": 572
    },
    {
      "epoch": 0.00573,
      "grad_norm": 0.8645205310284692,
      "learning_rate": 0.0017189999999999998,
      "loss": 5.1232,
      "step": 573
    },
    {
      "epoch": 0.00574,
      "grad_norm": 0.8122273817042681,
      "learning_rate": 0.001722,
      "loss": 5.1281,
      "step": 574
    },
    {
      "epoch": 0.00575,
      "grad_norm": 0.9590435469565912,
      "learning_rate": 0.001725,
      "loss": 5.1128,
      "step": 575
    },
    {
      "epoch": 0.00576,
      "grad_norm": 1.251082350842663,
      "learning_rate": 0.001728,
      "loss": 5.1225,
      "step": 576
    },
    {
      "epoch": 0.00577,
      "grad_norm": 0.7445616334136517,
      "learning_rate": 0.001731,
      "loss": 5.13,
      "step": 577
    },
    {
      "epoch": 0.00578,
      "grad_norm": 0.7845996621971332,
      "learning_rate": 0.0017339999999999999,
      "loss": 5.1117,
      "step": 578
    },
    {
      "epoch": 0.00579,
      "grad_norm": 0.886365355656363,
      "learning_rate": 0.0017369999999999998,
      "loss": 5.1037,
      "step": 579
    },
    {
      "epoch": 0.0058,
      "grad_norm": 0.8821301019883174,
      "learning_rate": 0.00174,
      "loss": 5.1259,
      "step": 580
    },
    {
      "epoch": 0.00581,
      "grad_norm": 0.7745940006716965,
      "learning_rate": 0.001743,
      "loss": 5.1161,
      "step": 581
    },
    {
      "epoch": 0.00582,
      "grad_norm": 0.8386253624392527,
      "learning_rate": 0.001746,
      "loss": 5.0979,
      "step": 582
    },
    {
      "epoch": 0.00583,
      "grad_norm": 0.8978101980978702,
      "learning_rate": 0.001749,
      "loss": 5.1236,
      "step": 583
    },
    {
      "epoch": 0.00584,
      "grad_norm": 0.9625436120988844,
      "learning_rate": 0.0017519999999999999,
      "loss": 5.1221,
      "step": 584
    },
    {
      "epoch": 0.00585,
      "grad_norm": 0.9007332505147065,
      "learning_rate": 0.0017549999999999998,
      "loss": 5.1079,
      "step": 585
    },
    {
      "epoch": 0.00586,
      "grad_norm": 0.8448788460460928,
      "learning_rate": 0.001758,
      "loss": 5.1127,
      "step": 586
    },
    {
      "epoch": 0.00587,
      "grad_norm": 0.8070040540273229,
      "learning_rate": 0.001761,
      "loss": 5.0903,
      "step": 587
    },
    {
      "epoch": 0.00588,
      "grad_norm": 0.8623574245532327,
      "learning_rate": 0.001764,
      "loss": 5.1032,
      "step": 588
    },
    {
      "epoch": 0.00589,
      "grad_norm": 0.8788484382751248,
      "learning_rate": 0.001767,
      "loss": 5.1023,
      "step": 589
    },
    {
      "epoch": 0.0059,
      "grad_norm": 0.88888081120713,
      "learning_rate": 0.0017699999999999999,
      "loss": 5.0951,
      "step": 590
    },
    {
      "epoch": 0.00591,
      "grad_norm": 0.8051815272157371,
      "learning_rate": 0.001773,
      "loss": 5.0807,
      "step": 591
    },
    {
      "epoch": 0.00592,
      "grad_norm": 0.9664766552178456,
      "learning_rate": 0.001776,
      "loss": 5.1164,
      "step": 592
    },
    {
      "epoch": 0.00593,
      "grad_norm": 0.933634159550007,
      "learning_rate": 0.001779,
      "loss": 5.1154,
      "step": 593
    },
    {
      "epoch": 0.00594,
      "grad_norm": 0.8388450047288221,
      "learning_rate": 0.001782,
      "loss": 5.1078,
      "step": 594
    },
    {
      "epoch": 0.00595,
      "grad_norm": 0.7870593334930018,
      "learning_rate": 0.001785,
      "loss": 5.0959,
      "step": 595
    },
    {
      "epoch": 0.00596,
      "grad_norm": 0.7663924438117601,
      "learning_rate": 0.0017879999999999999,
      "loss": 5.0923,
      "step": 596
    },
    {
      "epoch": 0.00597,
      "grad_norm": 0.9656629603017429,
      "learning_rate": 0.001791,
      "loss": 5.0773,
      "step": 597
    },
    {
      "epoch": 0.00598,
      "grad_norm": 0.9080147636693349,
      "learning_rate": 0.001794,
      "loss": 5.0893,
      "step": 598
    },
    {
      "epoch": 0.00599,
      "grad_norm": 0.732143169922112,
      "learning_rate": 0.001797,
      "loss": 5.0865,
      "step": 599
    },
    {
      "epoch": 0.006,
      "grad_norm": 0.5760572673229344,
      "learning_rate": 0.0018,
      "loss": 5.074,
      "step": 600
    },
    {
      "epoch": 0.00601,
      "grad_norm": 0.6602199941282026,
      "learning_rate": 0.001803,
      "loss": 5.0693,
      "step": 601
    },
    {
      "epoch": 0.00602,
      "grad_norm": 0.7069624755412358,
      "learning_rate": 0.0018059999999999999,
      "loss": 5.057,
      "step": 602
    },
    {
      "epoch": 0.00603,
      "grad_norm": 0.7134743776208471,
      "learning_rate": 0.001809,
      "loss": 5.0779,
      "step": 603
    },
    {
      "epoch": 0.00604,
      "grad_norm": 0.7455726311972387,
      "learning_rate": 0.001812,
      "loss": 5.0546,
      "step": 604
    },
    {
      "epoch": 0.00605,
      "grad_norm": 0.9135484051416195,
      "learning_rate": 0.001815,
      "loss": 5.0709,
      "step": 605
    },
    {
      "epoch": 0.00606,
      "grad_norm": 1.0710981560183195,
      "learning_rate": 0.001818,
      "loss": 5.0777,
      "step": 606
    },
    {
      "epoch": 0.00607,
      "grad_norm": 0.8059186248325771,
      "learning_rate": 0.001821,
      "loss": 5.0728,
      "step": 607
    },
    {
      "epoch": 0.00608,
      "grad_norm": 0.7260370848471782,
      "learning_rate": 0.001824,
      "loss": 5.0553,
      "step": 608
    },
    {
      "epoch": 0.00609,
      "grad_norm": 0.6501751004175685,
      "learning_rate": 0.001827,
      "loss": 5.0597,
      "step": 609
    },
    {
      "epoch": 0.0061,
      "grad_norm": 0.6148399666740725,
      "learning_rate": 0.00183,
      "loss": 5.0329,
      "step": 610
    },
    {
      "epoch": 0.00611,
      "grad_norm": 0.577060811762949,
      "learning_rate": 0.001833,
      "loss": 5.0322,
      "step": 611
    },
    {
      "epoch": 0.00612,
      "grad_norm": 0.6173055277138813,
      "learning_rate": 0.001836,
      "loss": 5.0402,
      "step": 612
    },
    {
      "epoch": 0.00613,
      "grad_norm": 0.615549230626737,
      "learning_rate": 0.001839,
      "loss": 5.0293,
      "step": 613
    },
    {
      "epoch": 0.00614,
      "grad_norm": 0.5145573627594662,
      "learning_rate": 0.001842,
      "loss": 5.0484,
      "step": 614
    },
    {
      "epoch": 0.00615,
      "grad_norm": 0.4622291494997757,
      "learning_rate": 0.001845,
      "loss": 5.0334,
      "step": 615
    },
    {
      "epoch": 0.00616,
      "grad_norm": 0.4156130098379491,
      "learning_rate": 0.001848,
      "loss": 5.0166,
      "step": 616
    },
    {
      "epoch": 0.00617,
      "grad_norm": 0.40499819493514183,
      "learning_rate": 0.001851,
      "loss": 5.0285,
      "step": 617
    },
    {
      "epoch": 0.00618,
      "grad_norm": 0.3547440364864465,
      "learning_rate": 0.001854,
      "loss": 5.0137,
      "step": 618
    },
    {
      "epoch": 0.00619,
      "grad_norm": 0.36790407868650016,
      "learning_rate": 0.001857,
      "loss": 5.0167,
      "step": 619
    },
    {
      "epoch": 0.0062,
      "grad_norm": 0.3873903126111541,
      "learning_rate": 0.00186,
      "loss": 5.0022,
      "step": 620
    },
    {
      "epoch": 0.00621,
      "grad_norm": 0.44545310636639085,
      "learning_rate": 0.001863,
      "loss": 5.0066,
      "step": 621
    },
    {
      "epoch": 0.00622,
      "grad_norm": 0.5872404207208428,
      "learning_rate": 0.001866,
      "loss": 5.0041,
      "step": 622
    },
    {
      "epoch": 0.00623,
      "grad_norm": 0.8188217453795142,
      "learning_rate": 0.001869,
      "loss": 4.9995,
      "step": 623
    },
    {
      "epoch": 0.00624,
      "grad_norm": 0.8582006227624419,
      "learning_rate": 0.001872,
      "loss": 5.0174,
      "step": 624
    },
    {
      "epoch": 0.00625,
      "grad_norm": 0.6624984450187602,
      "learning_rate": 0.001875,
      "loss": 5.0237,
      "step": 625
    },
    {
      "epoch": 0.00626,
      "grad_norm": 0.871266869538683,
      "learning_rate": 0.0018780000000000001,
      "loss": 5.0057,
      "step": 626
    },
    {
      "epoch": 0.00627,
      "grad_norm": 1.0014549660549559,
      "learning_rate": 0.001881,
      "loss": 5.0166,
      "step": 627
    },
    {
      "epoch": 0.00628,
      "grad_norm": 1.1152158811911206,
      "learning_rate": 0.001884,
      "loss": 5.0595,
      "step": 628
    },
    {
      "epoch": 0.00629,
      "grad_norm": 1.140057506672223,
      "learning_rate": 0.001887,
      "loss": 5.0271,
      "step": 629
    },
    {
      "epoch": 0.0063,
      "grad_norm": 1.1555923456286916,
      "learning_rate": 0.00189,
      "loss": 5.0294,
      "step": 630
    },
    {
      "epoch": 0.00631,
      "grad_norm": 1.0725579163260357,
      "learning_rate": 0.0018930000000000002,
      "loss": 5.0252,
      "step": 631
    },
    {
      "epoch": 0.00632,
      "grad_norm": 1.0622262545784495,
      "learning_rate": 0.0018960000000000001,
      "loss": 5.0506,
      "step": 632
    },
    {
      "epoch": 0.00633,
      "grad_norm": 1.1849469330653912,
      "learning_rate": 0.001899,
      "loss": 5.0591,
      "step": 633
    },
    {
      "epoch": 0.00634,
      "grad_norm": 1.186181168414293,
      "learning_rate": 0.001902,
      "loss": 5.0537,
      "step": 634
    },
    {
      "epoch": 0.00635,
      "grad_norm": 1.011082200052449,
      "learning_rate": 0.001905,
      "loss": 5.0328,
      "step": 635
    },
    {
      "epoch": 0.00636,
      "grad_norm": 1.1328438325981298,
      "learning_rate": 0.001908,
      "loss": 5.0454,
      "step": 636
    },
    {
      "epoch": 0.00637,
      "grad_norm": 0.9828095701538521,
      "learning_rate": 0.0019110000000000002,
      "loss": 5.0558,
      "step": 637
    },
    {
      "epoch": 0.00638,
      "grad_norm": 1.1150894310825157,
      "learning_rate": 0.0019140000000000001,
      "loss": 5.0551,
      "step": 638
    },
    {
      "epoch": 0.00639,
      "grad_norm": 0.9702746164579431,
      "learning_rate": 0.001917,
      "loss": 5.026,
      "step": 639
    },
    {
      "epoch": 0.0064,
      "grad_norm": 0.9463541889835033,
      "learning_rate": 0.00192,
      "loss": 5.0338,
      "step": 640
    },
    {
      "epoch": 0.00641,
      "grad_norm": 0.9922736299635262,
      "learning_rate": 0.001923,
      "loss": 5.0439,
      "step": 641
    },
    {
      "epoch": 0.00642,
      "grad_norm": 1.2833852757783575,
      "learning_rate": 0.001926,
      "loss": 5.0503,
      "step": 642
    },
    {
      "epoch": 0.00643,
      "grad_norm": 0.955455086522239,
      "learning_rate": 0.0019290000000000002,
      "loss": 5.0565,
      "step": 643
    },
    {
      "epoch": 0.00644,
      "grad_norm": 0.9001823386797868,
      "learning_rate": 0.0019320000000000001,
      "loss": 5.0314,
      "step": 644
    },
    {
      "epoch": 0.00645,
      "grad_norm": 0.7364586500805348,
      "learning_rate": 0.001935,
      "loss": 5.0304,
      "step": 645
    },
    {
      "epoch": 0.00646,
      "grad_norm": 0.6493450233680076,
      "learning_rate": 0.001938,
      "loss": 5.0293,
      "step": 646
    },
    {
      "epoch": 0.00647,
      "grad_norm": 0.6779095336835834,
      "learning_rate": 0.001941,
      "loss": 5.0401,
      "step": 647
    },
    {
      "epoch": 0.00648,
      "grad_norm": 0.6955209345108658,
      "learning_rate": 0.0019440000000000002,
      "loss": 5.0218,
      "step": 648
    },
    {
      "epoch": 0.00649,
      "grad_norm": 0.8304779062701974,
      "learning_rate": 0.0019470000000000002,
      "loss": 5.0072,
      "step": 649
    },
    {
      "epoch": 0.0065,
      "grad_norm": 0.820697007677232,
      "learning_rate": 0.0019500000000000001,
      "loss": 5.0315,
      "step": 650
    },
    {
      "epoch": 0.00651,
      "grad_norm": 0.7961110649030418,
      "learning_rate": 0.001953,
      "loss": 5.0121,
      "step": 651
    },
    {
      "epoch": 0.00652,
      "grad_norm": 0.8582193993973842,
      "learning_rate": 0.0019560000000000003,
      "loss": 5.009,
      "step": 652
    },
    {
      "epoch": 0.00653,
      "grad_norm": 0.8560401501333912,
      "learning_rate": 0.0019590000000000002,
      "loss": 5.0017,
      "step": 653
    },
    {
      "epoch": 0.00654,
      "grad_norm": 0.7831167086219499,
      "learning_rate": 0.001962,
      "loss": 4.9889,
      "step": 654
    },
    {
      "epoch": 0.00655,
      "grad_norm": 0.664585881803506,
      "learning_rate": 0.001965,
      "loss": 4.9976,
      "step": 655
    },
    {
      "epoch": 0.00656,
      "grad_norm": 0.6054298409810266,
      "learning_rate": 0.001968,
      "loss": 4.9857,
      "step": 656
    },
    {
      "epoch": 0.00657,
      "grad_norm": 0.557461380803133,
      "learning_rate": 0.001971,
      "loss": 4.9741,
      "step": 657
    },
    {
      "epoch": 0.00658,
      "grad_norm": 0.5329806623877416,
      "learning_rate": 0.001974,
      "loss": 4.9967,
      "step": 658
    },
    {
      "epoch": 0.00659,
      "grad_norm": 0.4733014627721161,
      "learning_rate": 0.001977,
      "loss": 4.9736,
      "step": 659
    },
    {
      "epoch": 0.0066,
      "grad_norm": 0.499559117562219,
      "learning_rate": 0.00198,
      "loss": 4.966,
      "step": 660
    },
    {
      "epoch": 0.00661,
      "grad_norm": 0.6099291254035663,
      "learning_rate": 0.001983,
      "loss": 4.9526,
      "step": 661
    },
    {
      "epoch": 0.00662,
      "grad_norm": 0.541377627450273,
      "learning_rate": 0.0019860000000000004,
      "loss": 4.9481,
      "step": 662
    },
    {
      "epoch": 0.00663,
      "grad_norm": 0.6125285838075794,
      "learning_rate": 0.0019890000000000003,
      "loss": 4.9658,
      "step": 663
    },
    {
      "epoch": 0.00664,
      "grad_norm": 0.7753721713708359,
      "learning_rate": 0.0019920000000000003,
      "loss": 4.9533,
      "step": 664
    },
    {
      "epoch": 0.00665,
      "grad_norm": 0.8561210143900951,
      "learning_rate": 0.0019950000000000002,
      "loss": 4.9333,
      "step": 665
    },
    {
      "epoch": 0.00666,
      "grad_norm": 0.7787737897544907,
      "learning_rate": 0.001998,
      "loss": 4.9543,
      "step": 666
    },
    {
      "epoch": 0.00667,
      "grad_norm": 0.693798446349636,
      "learning_rate": 0.002001,
      "loss": 4.9492,
      "step": 667
    },
    {
      "epoch": 0.00668,
      "grad_norm": 0.7363706742156494,
      "learning_rate": 0.002004,
      "loss": 4.9723,
      "step": 668
    },
    {
      "epoch": 0.00669,
      "grad_norm": 0.6301530155716453,
      "learning_rate": 0.002007,
      "loss": 4.9625,
      "step": 669
    },
    {
      "epoch": 0.0067,
      "grad_norm": 0.6196899316310271,
      "learning_rate": 0.00201,
      "loss": 4.9551,
      "step": 670
    },
    {
      "epoch": 0.00671,
      "grad_norm": 0.694548381041035,
      "learning_rate": 0.002013,
      "loss": 4.933,
      "step": 671
    },
    {
      "epoch": 0.00672,
      "grad_norm": 0.7621465658548545,
      "learning_rate": 0.002016,
      "loss": 4.9421,
      "step": 672
    },
    {
      "epoch": 0.00673,
      "grad_norm": 0.8075757173752571,
      "learning_rate": 0.002019,
      "loss": 4.9422,
      "step": 673
    },
    {
      "epoch": 0.00674,
      "grad_norm": 0.7418881135809064,
      "learning_rate": 0.0020220000000000004,
      "loss": 4.9464,
      "step": 674
    },
    {
      "epoch": 0.00675,
      "grad_norm": 0.6370116878967934,
      "learning_rate": 0.0020250000000000003,
      "loss": 4.9429,
      "step": 675
    },
    {
      "epoch": 0.00676,
      "grad_norm": 0.5146902497615841,
      "learning_rate": 0.0020280000000000003,
      "loss": 4.9317,
      "step": 676
    },
    {
      "epoch": 0.00677,
      "grad_norm": 0.5957426329332375,
      "learning_rate": 0.0020310000000000003,
      "loss": 4.9389,
      "step": 677
    },
    {
      "epoch": 0.00678,
      "grad_norm": 0.6059714528540561,
      "learning_rate": 0.0020340000000000002,
      "loss": 4.9473,
      "step": 678
    },
    {
      "epoch": 0.00679,
      "grad_norm": 0.6120329195118005,
      "learning_rate": 0.002037,
      "loss": 4.926,
      "step": 679
    },
    {
      "epoch": 0.0068,
      "grad_norm": 0.67013121421428,
      "learning_rate": 0.00204,
      "loss": 4.931,
      "step": 680
    },
    {
      "epoch": 0.00681,
      "grad_norm": 0.6629355745717784,
      "learning_rate": 0.002043,
      "loss": 4.9257,
      "step": 681
    },
    {
      "epoch": 0.00682,
      "grad_norm": 0.6203412878179155,
      "learning_rate": 0.002046,
      "loss": 4.9288,
      "step": 682
    },
    {
      "epoch": 0.00683,
      "grad_norm": 0.6526907123945536,
      "learning_rate": 0.002049,
      "loss": 4.911,
      "step": 683
    },
    {
      "epoch": 0.00684,
      "grad_norm": 0.7763655080494695,
      "learning_rate": 0.002052,
      "loss": 4.927,
      "step": 684
    },
    {
      "epoch": 0.00685,
      "grad_norm": 0.8186051850168985,
      "learning_rate": 0.0020550000000000004,
      "loss": 4.9269,
      "step": 685
    },
    {
      "epoch": 0.00686,
      "grad_norm": 0.7902721695493929,
      "learning_rate": 0.0020580000000000004,
      "loss": 4.9185,
      "step": 686
    },
    {
      "epoch": 0.00687,
      "grad_norm": 0.7114814619711178,
      "learning_rate": 0.0020610000000000003,
      "loss": 4.9075,
      "step": 687
    },
    {
      "epoch": 0.00688,
      "grad_norm": 0.6920604052885636,
      "learning_rate": 0.002064,
      "loss": 4.9133,
      "step": 688
    },
    {
      "epoch": 0.00689,
      "grad_norm": 0.757153733582835,
      "learning_rate": 0.002067,
      "loss": 4.9107,
      "step": 689
    },
    {
      "epoch": 0.0069,
      "grad_norm": 0.9661677354997352,
      "learning_rate": 0.00207,
      "loss": 4.9206,
      "step": 690
    },
    {
      "epoch": 0.00691,
      "grad_norm": 1.0541003703880292,
      "learning_rate": 0.0020729999999999998,
      "loss": 4.9045,
      "step": 691
    },
    {
      "epoch": 0.00692,
      "grad_norm": 0.8794573104042516,
      "learning_rate": 0.0020759999999999997,
      "loss": 4.923,
      "step": 692
    },
    {
      "epoch": 0.00693,
      "grad_norm": 0.9176864844747288,
      "learning_rate": 0.0020789999999999997,
      "loss": 4.9239,
      "step": 693
    },
    {
      "epoch": 0.00694,
      "grad_norm": 0.9209969930842488,
      "learning_rate": 0.002082,
      "loss": 4.9074,
      "step": 694
    },
    {
      "epoch": 0.00695,
      "grad_norm": 1.1695252405572139,
      "learning_rate": 0.002085,
      "loss": 4.9261,
      "step": 695
    },
    {
      "epoch": 0.00696,
      "grad_norm": 0.9145612208618356,
      "learning_rate": 0.002088,
      "loss": 4.919,
      "step": 696
    },
    {
      "epoch": 0.00697,
      "grad_norm": 1.1036978009332064,
      "learning_rate": 0.002091,
      "loss": 4.9333,
      "step": 697
    },
    {
      "epoch": 0.00698,
      "grad_norm": 1.0287653208524956,
      "learning_rate": 0.002094,
      "loss": 4.9255,
      "step": 698
    },
    {
      "epoch": 0.00699,
      "grad_norm": 0.9310985158925286,
      "learning_rate": 0.002097,
      "loss": 4.9194,
      "step": 699
    },
    {
      "epoch": 0.007,
      "grad_norm": 0.917395048546633,
      "learning_rate": 0.0021,
      "loss": 4.9296,
      "step": 700
    },
    {
      "epoch": 0.00701,
      "grad_norm": 0.834332437065761,
      "learning_rate": 0.002103,
      "loss": 4.9089,
      "step": 701
    },
    {
      "epoch": 0.00702,
      "grad_norm": 0.8281083071877697,
      "learning_rate": 0.002106,
      "loss": 4.9094,
      "step": 702
    },
    {
      "epoch": 0.00703,
      "grad_norm": 0.8445695596642441,
      "learning_rate": 0.0021089999999999998,
      "loss": 4.8989,
      "step": 703
    },
    {
      "epoch": 0.00704,
      "grad_norm": 0.7492253201313137,
      "learning_rate": 0.0021119999999999997,
      "loss": 4.8989,
      "step": 704
    },
    {
      "epoch": 0.00705,
      "grad_norm": 0.735879432986631,
      "learning_rate": 0.002115,
      "loss": 4.9035,
      "step": 705
    },
    {
      "epoch": 0.00706,
      "grad_norm": 0.6973072916647776,
      "learning_rate": 0.002118,
      "loss": 4.9057,
      "step": 706
    },
    {
      "epoch": 0.00707,
      "grad_norm": 0.629484514979941,
      "learning_rate": 0.002121,
      "loss": 4.8808,
      "step": 707
    },
    {
      "epoch": 0.00708,
      "grad_norm": 0.5238472656698838,
      "learning_rate": 0.002124,
      "loss": 4.8573,
      "step": 708
    },
    {
      "epoch": 0.00709,
      "grad_norm": 0.5712907279215416,
      "learning_rate": 0.002127,
      "loss": 4.889,
      "step": 709
    },
    {
      "epoch": 0.0071,
      "grad_norm": 0.7552376956739056,
      "learning_rate": 0.00213,
      "loss": 4.8786,
      "step": 710
    },
    {
      "epoch": 0.00711,
      "grad_norm": 0.9238310631482564,
      "learning_rate": 0.002133,
      "loss": 4.8637,
      "step": 711
    },
    {
      "epoch": 0.00712,
      "grad_norm": 0.9459038147293442,
      "learning_rate": 0.002136,
      "loss": 4.8869,
      "step": 712
    },
    {
      "epoch": 0.00713,
      "grad_norm": 0.960196647997872,
      "learning_rate": 0.002139,
      "loss": 4.9055,
      "step": 713
    },
    {
      "epoch": 0.00714,
      "grad_norm": 0.9883902427962058,
      "learning_rate": 0.002142,
      "loss": 4.9036,
      "step": 714
    },
    {
      "epoch": 0.00715,
      "grad_norm": 1.091153020847237,
      "learning_rate": 0.0021449999999999998,
      "loss": 4.8821,
      "step": 715
    },
    {
      "epoch": 0.00716,
      "grad_norm": 0.9092241477742985,
      "learning_rate": 0.002148,
      "loss": 4.8679,
      "step": 716
    },
    {
      "epoch": 0.00717,
      "grad_norm": 0.9933217219488532,
      "learning_rate": 0.002151,
      "loss": 4.8972,
      "step": 717
    },
    {
      "epoch": 0.00718,
      "grad_norm": 1.0107964436698889,
      "learning_rate": 0.002154,
      "loss": 4.8709,
      "step": 718
    },
    {
      "epoch": 0.00719,
      "grad_norm": 0.9093998544086449,
      "learning_rate": 0.002157,
      "loss": 4.8869,
      "step": 719
    },
    {
      "epoch": 0.0072,
      "grad_norm": 0.8253722879391543,
      "learning_rate": 0.00216,
      "loss": 4.8711,
      "step": 720
    },
    {
      "epoch": 0.00721,
      "grad_norm": 0.7950204631220138,
      "learning_rate": 0.002163,
      "loss": 4.8781,
      "step": 721
    },
    {
      "epoch": 0.00722,
      "grad_norm": 0.7616416140294759,
      "learning_rate": 0.002166,
      "loss": 4.8517,
      "step": 722
    },
    {
      "epoch": 0.00723,
      "grad_norm": 0.8494335965132823,
      "learning_rate": 0.002169,
      "loss": 4.8626,
      "step": 723
    },
    {
      "epoch": 0.00724,
      "grad_norm": 0.9441673028016423,
      "learning_rate": 0.002172,
      "loss": 4.8621,
      "step": 724
    },
    {
      "epoch": 0.00725,
      "grad_norm": 0.8751268950403425,
      "learning_rate": 0.002175,
      "loss": 4.8595,
      "step": 725
    },
    {
      "epoch": 0.00726,
      "grad_norm": 0.9089850912837197,
      "learning_rate": 0.002178,
      "loss": 4.8628,
      "step": 726
    },
    {
      "epoch": 0.00727,
      "grad_norm": 0.8888908119178632,
      "learning_rate": 0.0021809999999999998,
      "loss": 4.8505,
      "step": 727
    },
    {
      "epoch": 0.00728,
      "grad_norm": 0.9736770183257887,
      "learning_rate": 0.002184,
      "loss": 4.8689,
      "step": 728
    },
    {
      "epoch": 0.00729,
      "grad_norm": 0.8001407133698992,
      "learning_rate": 0.002187,
      "loss": 4.8549,
      "step": 729
    },
    {
      "epoch": 0.0073,
      "grad_norm": 0.6821558897279537,
      "learning_rate": 0.00219,
      "loss": 4.8548,
      "step": 730
    },
    {
      "epoch": 0.00731,
      "grad_norm": 0.6831491892882273,
      "learning_rate": 0.002193,
      "loss": 4.8443,
      "step": 731
    },
    {
      "epoch": 0.00732,
      "grad_norm": 0.7429226676106812,
      "learning_rate": 0.002196,
      "loss": 4.8496,
      "step": 732
    },
    {
      "epoch": 0.00733,
      "grad_norm": 0.7471373337600634,
      "learning_rate": 0.002199,
      "loss": 4.8306,
      "step": 733
    },
    {
      "epoch": 0.00734,
      "grad_norm": 0.7160425997494153,
      "learning_rate": 0.002202,
      "loss": 4.845,
      "step": 734
    },
    {
      "epoch": 0.00735,
      "grad_norm": 0.657968205956134,
      "learning_rate": 0.002205,
      "loss": 4.8539,
      "step": 735
    },
    {
      "epoch": 0.00736,
      "grad_norm": 0.7318419955485544,
      "learning_rate": 0.002208,
      "loss": 4.8371,
      "step": 736
    },
    {
      "epoch": 0.00737,
      "grad_norm": 0.8187387102115173,
      "learning_rate": 0.002211,
      "loss": 4.8083,
      "step": 737
    },
    {
      "epoch": 0.00738,
      "grad_norm": 1.0775553672977198,
      "learning_rate": 0.002214,
      "loss": 4.8215,
      "step": 738
    },
    {
      "epoch": 0.00739,
      "grad_norm": 1.0127446069090642,
      "learning_rate": 0.0022170000000000002,
      "loss": 4.8211,
      "step": 739
    },
    {
      "epoch": 0.0074,
      "grad_norm": 0.8746324375476564,
      "learning_rate": 0.00222,
      "loss": 4.8143,
      "step": 740
    },
    {
      "epoch": 0.00741,
      "grad_norm": 1.0041320632210426,
      "learning_rate": 0.002223,
      "loss": 4.7944,
      "step": 741
    },
    {
      "epoch": 0.00742,
      "grad_norm": 1.1063519275220866,
      "learning_rate": 0.002226,
      "loss": 4.836,
      "step": 742
    },
    {
      "epoch": 0.00743,
      "grad_norm": 0.9221104447659358,
      "learning_rate": 0.002229,
      "loss": 4.8024,
      "step": 743
    },
    {
      "epoch": 0.00744,
      "grad_norm": 0.9424830767002879,
      "learning_rate": 0.002232,
      "loss": 4.8123,
      "step": 744
    },
    {
      "epoch": 0.00745,
      "grad_norm": 1.0188289984750667,
      "learning_rate": 0.002235,
      "loss": 4.8151,
      "step": 745
    },
    {
      "epoch": 0.00746,
      "grad_norm": 1.0287722855005268,
      "learning_rate": 0.002238,
      "loss": 4.8238,
      "step": 746
    },
    {
      "epoch": 0.00747,
      "grad_norm": 1.0487862769229943,
      "learning_rate": 0.002241,
      "loss": 4.8099,
      "step": 747
    },
    {
      "epoch": 0.00748,
      "grad_norm": 0.8549676783684295,
      "learning_rate": 0.002244,
      "loss": 4.8084,
      "step": 748
    },
    {
      "epoch": 0.00749,
      "grad_norm": 0.8567053944604911,
      "learning_rate": 0.002247,
      "loss": 4.787,
      "step": 749
    },
    {
      "epoch": 0.0075,
      "grad_norm": 1.0511964127667834,
      "learning_rate": 0.0022500000000000003,
      "loss": 4.8309,
      "step": 750
    },
    {
      "epoch": 0.00751,
      "grad_norm": 0.9720966734602534,
      "learning_rate": 0.0022530000000000002,
      "loss": 4.8211,
      "step": 751
    },
    {
      "epoch": 0.00752,
      "grad_norm": 0.8427538199759678,
      "learning_rate": 0.002256,
      "loss": 4.777,
      "step": 752
    },
    {
      "epoch": 0.00753,
      "grad_norm": 1.0504906636895608,
      "learning_rate": 0.002259,
      "loss": 4.8189,
      "step": 753
    },
    {
      "epoch": 0.00754,
      "grad_norm": 0.9387358479584846,
      "learning_rate": 0.002262,
      "loss": 4.8364,
      "step": 754
    },
    {
      "epoch": 0.00755,
      "grad_norm": 0.7913656975262588,
      "learning_rate": 0.002265,
      "loss": 4.7934,
      "step": 755
    },
    {
      "epoch": 0.00756,
      "grad_norm": 0.7692929115953351,
      "learning_rate": 0.002268,
      "loss": 4.7756,
      "step": 756
    },
    {
      "epoch": 0.00757,
      "grad_norm": 0.7739955836145438,
      "learning_rate": 0.002271,
      "loss": 4.7882,
      "step": 757
    },
    {
      "epoch": 0.00758,
      "grad_norm": 0.8110040425710608,
      "learning_rate": 0.002274,
      "loss": 4.7648,
      "step": 758
    },
    {
      "epoch": 0.00759,
      "grad_norm": 0.8073791732331314,
      "learning_rate": 0.002277,
      "loss": 4.785,
      "step": 759
    },
    {
      "epoch": 0.0076,
      "grad_norm": 0.7576731140462794,
      "learning_rate": 0.00228,
      "loss": 4.7713,
      "step": 760
    },
    {
      "epoch": 0.00761,
      "grad_norm": 0.749632110976692,
      "learning_rate": 0.002283,
      "loss": 4.7377,
      "step": 761
    },
    {
      "epoch": 0.00762,
      "grad_norm": 0.7853246221677522,
      "learning_rate": 0.0022860000000000003,
      "loss": 4.7232,
      "step": 762
    },
    {
      "epoch": 0.00763,
      "grad_norm": 0.7445635461156005,
      "learning_rate": 0.0022890000000000002,
      "loss": 4.7582,
      "step": 763
    },
    {
      "epoch": 0.00764,
      "grad_norm": 0.7229992019662822,
      "learning_rate": 0.002292,
      "loss": 4.761,
      "step": 764
    },
    {
      "epoch": 0.00765,
      "grad_norm": 0.753778651079225,
      "learning_rate": 0.002295,
      "loss": 4.748,
      "step": 765
    },
    {
      "epoch": 0.00766,
      "grad_norm": 0.6446844137748907,
      "learning_rate": 0.002298,
      "loss": 4.7414,
      "step": 766
    },
    {
      "epoch": 0.00767,
      "grad_norm": 0.610956287563313,
      "learning_rate": 0.002301,
      "loss": 4.7382,
      "step": 767
    },
    {
      "epoch": 0.00768,
      "grad_norm": 0.5534988019273127,
      "learning_rate": 0.002304,
      "loss": 4.7365,
      "step": 768
    },
    {
      "epoch": 0.00769,
      "grad_norm": 0.6790892369300202,
      "learning_rate": 0.002307,
      "loss": 4.7418,
      "step": 769
    },
    {
      "epoch": 0.0077,
      "grad_norm": 0.8121255147206851,
      "learning_rate": 0.00231,
      "loss": 4.7084,
      "step": 770
    },
    {
      "epoch": 0.00771,
      "grad_norm": 0.860209240156232,
      "learning_rate": 0.002313,
      "loss": 4.7446,
      "step": 771
    },
    {
      "epoch": 0.00772,
      "grad_norm": 0.860546267364981,
      "learning_rate": 0.002316,
      "loss": 4.7297,
      "step": 772
    },
    {
      "epoch": 0.00773,
      "grad_norm": 0.8140139366688685,
      "learning_rate": 0.0023190000000000003,
      "loss": 4.7319,
      "step": 773
    },
    {
      "epoch": 0.00774,
      "grad_norm": 0.8843308751287595,
      "learning_rate": 0.0023220000000000003,
      "loss": 4.7474,
      "step": 774
    },
    {
      "epoch": 0.00775,
      "grad_norm": 0.8212427196933955,
      "learning_rate": 0.0023250000000000002,
      "loss": 4.7389,
      "step": 775
    },
    {
      "epoch": 0.00776,
      "grad_norm": 0.7590181685436337,
      "learning_rate": 0.002328,
      "loss": 4.7362,
      "step": 776
    },
    {
      "epoch": 0.00777,
      "grad_norm": 0.7163129834638192,
      "learning_rate": 0.002331,
      "loss": 4.711,
      "step": 777
    },
    {
      "epoch": 0.00778,
      "grad_norm": 0.5704385831566832,
      "learning_rate": 0.002334,
      "loss": 4.7139,
      "step": 778
    },
    {
      "epoch": 0.00779,
      "grad_norm": 0.6376822022096883,
      "learning_rate": 0.002337,
      "loss": 4.6765,
      "step": 779
    },
    {
      "epoch": 0.0078,
      "grad_norm": 0.6749602286292105,
      "learning_rate": 0.00234,
      "loss": 4.6934,
      "step": 780
    },
    {
      "epoch": 0.00781,
      "grad_norm": 0.7110957632967577,
      "learning_rate": 0.002343,
      "loss": 4.6953,
      "step": 781
    },
    {
      "epoch": 0.00782,
      "grad_norm": 0.6944526657973698,
      "learning_rate": 0.002346,
      "loss": 4.6887,
      "step": 782
    },
    {
      "epoch": 0.00783,
      "grad_norm": 0.7011179026115173,
      "learning_rate": 0.002349,
      "loss": 4.6685,
      "step": 783
    },
    {
      "epoch": 0.00784,
      "grad_norm": 0.7292703169217482,
      "learning_rate": 0.002352,
      "loss": 4.6842,
      "step": 784
    },
    {
      "epoch": 0.00785,
      "grad_norm": 0.6993372864201758,
      "learning_rate": 0.0023550000000000003,
      "loss": 4.6776,
      "step": 785
    },
    {
      "epoch": 0.00786,
      "grad_norm": 0.758602493265201,
      "learning_rate": 0.0023580000000000003,
      "loss": 4.6843,
      "step": 786
    },
    {
      "epoch": 0.00787,
      "grad_norm": 0.8960009188651508,
      "learning_rate": 0.0023610000000000003,
      "loss": 4.6883,
      "step": 787
    },
    {
      "epoch": 0.00788,
      "grad_norm": 0.9005115946196935,
      "learning_rate": 0.002364,
      "loss": 4.6894,
      "step": 788
    },
    {
      "epoch": 0.00789,
      "grad_norm": 0.9107967107186621,
      "learning_rate": 0.002367,
      "loss": 4.7167,
      "step": 789
    },
    {
      "epoch": 0.0079,
      "grad_norm": 0.8913303624885218,
      "learning_rate": 0.00237,
      "loss": 4.6905,
      "step": 790
    },
    {
      "epoch": 0.00791,
      "grad_norm": 0.9033565188226677,
      "learning_rate": 0.002373,
      "loss": 4.6536,
      "step": 791
    },
    {
      "epoch": 0.00792,
      "grad_norm": 0.876812764850389,
      "learning_rate": 0.002376,
      "loss": 4.6841,
      "step": 792
    },
    {
      "epoch": 0.00793,
      "grad_norm": 0.8853769972879193,
      "learning_rate": 0.002379,
      "loss": 4.6794,
      "step": 793
    },
    {
      "epoch": 0.00794,
      "grad_norm": 0.9106991776684247,
      "learning_rate": 0.002382,
      "loss": 4.6921,
      "step": 794
    },
    {
      "epoch": 0.00795,
      "grad_norm": 0.8863762739828056,
      "learning_rate": 0.002385,
      "loss": 4.6703,
      "step": 795
    },
    {
      "epoch": 0.00796,
      "grad_norm": 0.8699126455036565,
      "learning_rate": 0.0023880000000000004,
      "loss": 4.664,
      "step": 796
    },
    {
      "epoch": 0.00797,
      "grad_norm": 0.8543487171707993,
      "learning_rate": 0.0023910000000000003,
      "loss": 4.7077,
      "step": 797
    },
    {
      "epoch": 0.00798,
      "grad_norm": 0.8580039409853985,
      "learning_rate": 0.0023940000000000003,
      "loss": 4.6717,
      "step": 798
    },
    {
      "epoch": 0.00799,
      "grad_norm": 0.9172496139502885,
      "learning_rate": 0.0023970000000000003,
      "loss": 4.7059,
      "step": 799
    },
    {
      "epoch": 0.008,
      "grad_norm": 0.850755534424379,
      "learning_rate": 0.0024000000000000002,
      "loss": 4.6592,
      "step": 800
    },
    {
      "epoch": 0.00801,
      "grad_norm": 0.9527629416181698,
      "learning_rate": 0.002403,
      "loss": 4.6941,
      "step": 801
    },
    {
      "epoch": 0.00802,
      "grad_norm": 0.907546457744482,
      "learning_rate": 0.002406,
      "loss": 4.6836,
      "step": 802
    },
    {
      "epoch": 0.00803,
      "grad_norm": 0.8445766186911342,
      "learning_rate": 0.002409,
      "loss": 4.6598,
      "step": 803
    },
    {
      "epoch": 0.00804,
      "grad_norm": 0.8905553328902869,
      "learning_rate": 0.002412,
      "loss": 4.6598,
      "step": 804
    },
    {
      "epoch": 0.00805,
      "grad_norm": 0.8936903365864526,
      "learning_rate": 0.002415,
      "loss": 4.6624,
      "step": 805
    },
    {
      "epoch": 0.00806,
      "grad_norm": 0.8647204208273495,
      "learning_rate": 0.002418,
      "loss": 4.6755,
      "step": 806
    },
    {
      "epoch": 0.00807,
      "grad_norm": 0.8494760155293042,
      "learning_rate": 0.0024210000000000004,
      "loss": 4.6694,
      "step": 807
    },
    {
      "epoch": 0.00808,
      "grad_norm": 0.7376236468427481,
      "learning_rate": 0.0024240000000000004,
      "loss": 4.65,
      "step": 808
    },
    {
      "epoch": 0.00809,
      "grad_norm": 0.7397890518039199,
      "learning_rate": 0.0024270000000000003,
      "loss": 4.6449,
      "step": 809
    },
    {
      "epoch": 0.0081,
      "grad_norm": 0.6742672790291327,
      "learning_rate": 0.0024300000000000003,
      "loss": 4.6639,
      "step": 810
    },
    {
      "epoch": 0.00811,
      "grad_norm": 0.6344626039053871,
      "learning_rate": 0.0024330000000000003,
      "loss": 4.6372,
      "step": 811
    },
    {
      "epoch": 0.00812,
      "grad_norm": 0.6264235852250682,
      "learning_rate": 0.0024360000000000002,
      "loss": 4.6138,
      "step": 812
    },
    {
      "epoch": 0.00813,
      "grad_norm": 0.6033433029827218,
      "learning_rate": 0.0024389999999999998,
      "loss": 4.643,
      "step": 813
    },
    {
      "epoch": 0.00814,
      "grad_norm": 0.6217940219682172,
      "learning_rate": 0.0024419999999999997,
      "loss": 4.6512,
      "step": 814
    },
    {
      "epoch": 0.00815,
      "grad_norm": 0.6033575847456577,
      "learning_rate": 0.0024449999999999997,
      "loss": 4.6541,
      "step": 815
    },
    {
      "epoch": 0.00816,
      "grad_norm": 0.5100796560015926,
      "learning_rate": 0.002448,
      "loss": 4.6231,
      "step": 816
    },
    {
      "epoch": 0.00817,
      "grad_norm": 0.49585009337112085,
      "learning_rate": 0.002451,
      "loss": 4.6301,
      "step": 817
    },
    {
      "epoch": 0.00818,
      "grad_norm": 0.5266854681962923,
      "learning_rate": 0.002454,
      "loss": 4.6109,
      "step": 818
    },
    {
      "epoch": 0.00819,
      "grad_norm": 0.6082191743538553,
      "learning_rate": 0.002457,
      "loss": 4.6237,
      "step": 819
    },
    {
      "epoch": 0.0082,
      "grad_norm": 0.5983769962466982,
      "learning_rate": 0.00246,
      "loss": 4.6093,
      "step": 820
    },
    {
      "epoch": 0.00821,
      "grad_norm": 0.6452383077710647,
      "learning_rate": 0.002463,
      "loss": 4.5921,
      "step": 821
    },
    {
      "epoch": 0.00822,
      "grad_norm": 0.7904685465267983,
      "learning_rate": 0.002466,
      "loss": 4.5962,
      "step": 822
    },
    {
      "epoch": 0.00823,
      "grad_norm": 0.9334545600211804,
      "learning_rate": 0.002469,
      "loss": 4.6151,
      "step": 823
    },
    {
      "epoch": 0.00824,
      "grad_norm": 1.0173145160580905,
      "learning_rate": 0.002472,
      "loss": 4.6432,
      "step": 824
    },
    {
      "epoch": 0.00825,
      "grad_norm": 1.3805405563709947,
      "learning_rate": 0.0024749999999999998,
      "loss": 4.6688,
      "step": 825
    },
    {
      "epoch": 0.00826,
      "grad_norm": 0.8722736889569705,
      "learning_rate": 0.0024779999999999997,
      "loss": 4.608,
      "step": 826
    },
    {
      "epoch": 0.00827,
      "grad_norm": 0.9315441314299773,
      "learning_rate": 0.002481,
      "loss": 4.6415,
      "step": 827
    },
    {
      "epoch": 0.00828,
      "grad_norm": 1.0512546819946473,
      "learning_rate": 0.002484,
      "loss": 4.6408,
      "step": 828
    },
    {
      "epoch": 0.00829,
      "grad_norm": 1.0438218855109669,
      "learning_rate": 0.002487,
      "loss": 4.6506,
      "step": 829
    },
    {
      "epoch": 0.0083,
      "grad_norm": 0.9301285778060441,
      "learning_rate": 0.00249,
      "loss": 4.6477,
      "step": 830
    },
    {
      "epoch": 0.00831,
      "grad_norm": 0.9032044535603423,
      "learning_rate": 0.002493,
      "loss": 4.629,
      "step": 831
    },
    {
      "epoch": 0.00832,
      "grad_norm": 0.8730808748017225,
      "learning_rate": 0.002496,
      "loss": 4.6303,
      "step": 832
    },
    {
      "epoch": 0.00833,
      "grad_norm": 0.8092334117771479,
      "learning_rate": 0.002499,
      "loss": 4.6217,
      "step": 833
    },
    {
      "epoch": 0.00834,
      "grad_norm": 0.7831190344913049,
      "learning_rate": 0.002502,
      "loss": 4.624,
      "step": 834
    },
    {
      "epoch": 0.00835,
      "grad_norm": 0.6494366058082618,
      "learning_rate": 0.002505,
      "loss": 4.6082,
      "step": 835
    },
    {
      "epoch": 0.00836,
      "grad_norm": 0.6821091092072289,
      "learning_rate": 0.002508,
      "loss": 4.6348,
      "step": 836
    },
    {
      "epoch": 0.00837,
      "grad_norm": 0.7523253894726831,
      "learning_rate": 0.0025109999999999998,
      "loss": 4.5808,
      "step": 837
    },
    {
      "epoch": 0.00838,
      "grad_norm": 0.7788634101452747,
      "learning_rate": 0.0025139999999999997,
      "loss": 4.6077,
      "step": 838
    },
    {
      "epoch": 0.00839,
      "grad_norm": 0.8138659157639668,
      "learning_rate": 0.002517,
      "loss": 4.6083,
      "step": 839
    },
    {
      "epoch": 0.0084,
      "grad_norm": 0.8732162622457459,
      "learning_rate": 0.00252,
      "loss": 4.5955,
      "step": 840
    },
    {
      "epoch": 0.00841,
      "grad_norm": 1.019417984642289,
      "learning_rate": 0.002523,
      "loss": 4.598,
      "step": 841
    },
    {
      "epoch": 0.00842,
      "grad_norm": 0.9409472966208113,
      "learning_rate": 0.002526,
      "loss": 4.6049,
      "step": 842
    },
    {
      "epoch": 0.00843,
      "grad_norm": 0.9226761303774836,
      "learning_rate": 0.002529,
      "loss": 4.5993,
      "step": 843
    },
    {
      "epoch": 0.00844,
      "grad_norm": 0.9652698373224984,
      "learning_rate": 0.002532,
      "loss": 4.6327,
      "step": 844
    },
    {
      "epoch": 0.00845,
      "grad_norm": 0.9308804697564665,
      "learning_rate": 0.002535,
      "loss": 4.5933,
      "step": 845
    },
    {
      "epoch": 0.00846,
      "grad_norm": 0.9944671667950292,
      "learning_rate": 0.002538,
      "loss": 4.6213,
      "step": 846
    },
    {
      "epoch": 0.00847,
      "grad_norm": 0.7973628646086437,
      "learning_rate": 0.002541,
      "loss": 4.6012,
      "step": 847
    },
    {
      "epoch": 0.00848,
      "grad_norm": 0.7238143616386193,
      "learning_rate": 0.002544,
      "loss": 4.5901,
      "step": 848
    },
    {
      "epoch": 0.00849,
      "grad_norm": 0.8442325507863795,
      "learning_rate": 0.002547,
      "loss": 4.5938,
      "step": 849
    },
    {
      "epoch": 0.0085,
      "grad_norm": 0.7603726939752414,
      "learning_rate": 0.00255,
      "loss": 4.5976,
      "step": 850
    },
    {
      "epoch": 0.00851,
      "grad_norm": 0.8865277336062035,
      "learning_rate": 0.002553,
      "loss": 4.6202,
      "step": 851
    },
    {
      "epoch": 0.00852,
      "grad_norm": 0.9253666339112696,
      "learning_rate": 0.002556,
      "loss": 4.6122,
      "step": 852
    },
    {
      "epoch": 0.00853,
      "grad_norm": 0.9517307300698885,
      "learning_rate": 0.002559,
      "loss": 4.5919,
      "step": 853
    },
    {
      "epoch": 0.00854,
      "grad_norm": 1.0310567151805228,
      "learning_rate": 0.002562,
      "loss": 4.5922,
      "step": 854
    },
    {
      "epoch": 0.00855,
      "grad_norm": 1.154844172511891,
      "learning_rate": 0.002565,
      "loss": 4.5856,
      "step": 855
    },
    {
      "epoch": 0.00856,
      "grad_norm": 0.9664055608114693,
      "learning_rate": 0.002568,
      "loss": 4.6075,
      "step": 856
    },
    {
      "epoch": 0.00857,
      "grad_norm": 1.116045732107275,
      "learning_rate": 0.002571,
      "loss": 4.606,
      "step": 857
    },
    {
      "epoch": 0.00858,
      "grad_norm": 0.7861495756039047,
      "learning_rate": 0.002574,
      "loss": 4.5872,
      "step": 858
    },
    {
      "epoch": 0.00859,
      "grad_norm": 0.6214173325635484,
      "learning_rate": 0.002577,
      "loss": 4.5716,
      "step": 859
    },
    {
      "epoch": 0.0086,
      "grad_norm": 0.6164572973900683,
      "learning_rate": 0.00258,
      "loss": 4.5977,
      "step": 860
    },
    {
      "epoch": 0.00861,
      "grad_norm": 0.5506301368516001,
      "learning_rate": 0.0025830000000000002,
      "loss": 4.6069,
      "step": 861
    },
    {
      "epoch": 0.00862,
      "grad_norm": 0.601070156302512,
      "learning_rate": 0.002586,
      "loss": 4.5567,
      "step": 862
    },
    {
      "epoch": 0.00863,
      "grad_norm": 0.6954858119983386,
      "learning_rate": 0.002589,
      "loss": 4.58,
      "step": 863
    },
    {
      "epoch": 0.00864,
      "grad_norm": 0.697853484629387,
      "learning_rate": 0.002592,
      "loss": 4.533,
      "step": 864
    },
    {
      "epoch": 0.00865,
      "grad_norm": 0.6171533334128538,
      "learning_rate": 0.002595,
      "loss": 4.5507,
      "step": 865
    },
    {
      "epoch": 0.00866,
      "grad_norm": 0.6615774116930657,
      "learning_rate": 0.002598,
      "loss": 4.5725,
      "step": 866
    },
    {
      "epoch": 0.00867,
      "grad_norm": 0.7744523515971641,
      "learning_rate": 0.002601,
      "loss": 4.5651,
      "step": 867
    },
    {
      "epoch": 0.00868,
      "grad_norm": 0.820881730278172,
      "learning_rate": 0.002604,
      "loss": 4.5507,
      "step": 868
    },
    {
      "epoch": 0.00869,
      "grad_norm": 0.7803750245131083,
      "learning_rate": 0.002607,
      "loss": 4.5569,
      "step": 869
    },
    {
      "epoch": 0.0087,
      "grad_norm": 0.7384702018181832,
      "learning_rate": 0.00261,
      "loss": 4.5432,
      "step": 870
    },
    {
      "epoch": 0.00871,
      "grad_norm": 0.7238577244427861,
      "learning_rate": 0.002613,
      "loss": 4.5594,
      "step": 871
    },
    {
      "epoch": 0.00872,
      "grad_norm": 0.6596009767396808,
      "learning_rate": 0.002616,
      "loss": 4.5386,
      "step": 872
    },
    {
      "epoch": 0.00873,
      "grad_norm": 0.657018760483689,
      "learning_rate": 0.0026190000000000002,
      "loss": 4.5487,
      "step": 873
    },
    {
      "epoch": 0.00874,
      "grad_norm": 0.5096984570514209,
      "learning_rate": 0.002622,
      "loss": 4.5587,
      "step": 874
    },
    {
      "epoch": 0.00875,
      "grad_norm": 0.5917862001329391,
      "learning_rate": 0.002625,
      "loss": 4.509,
      "step": 875
    },
    {
      "epoch": 0.00876,
      "grad_norm": 0.6358204645109454,
      "learning_rate": 0.002628,
      "loss": 4.5421,
      "step": 876
    },
    {
      "epoch": 0.00877,
      "grad_norm": 0.622105352843196,
      "learning_rate": 0.002631,
      "loss": 4.5557,
      "step": 877
    },
    {
      "epoch": 0.00878,
      "grad_norm": 0.6489802918731719,
      "learning_rate": 0.002634,
      "loss": 4.4926,
      "step": 878
    },
    {
      "epoch": 0.00879,
      "grad_norm": 0.7198651946841444,
      "learning_rate": 0.002637,
      "loss": 4.5246,
      "step": 879
    },
    {
      "epoch": 0.0088,
      "grad_norm": 0.807778517531823,
      "learning_rate": 0.00264,
      "loss": 4.5381,
      "step": 880
    },
    {
      "epoch": 0.00881,
      "grad_norm": 0.6840558014472901,
      "learning_rate": 0.002643,
      "loss": 4.5363,
      "step": 881
    },
    {
      "epoch": 0.00882,
      "grad_norm": 0.6535717374503174,
      "learning_rate": 0.002646,
      "loss": 4.532,
      "step": 882
    },
    {
      "epoch": 0.00883,
      "grad_norm": 0.6598608882990905,
      "learning_rate": 0.002649,
      "loss": 4.528,
      "step": 883
    },
    {
      "epoch": 0.00884,
      "grad_norm": 0.5423501635938441,
      "learning_rate": 0.0026520000000000003,
      "loss": 4.5066,
      "step": 884
    },
    {
      "epoch": 0.00885,
      "grad_norm": 0.5323717529728587,
      "learning_rate": 0.0026550000000000002,
      "loss": 4.5028,
      "step": 885
    },
    {
      "epoch": 0.00886,
      "grad_norm": 0.6188094567817924,
      "learning_rate": 0.002658,
      "loss": 4.5234,
      "step": 886
    },
    {
      "epoch": 0.00887,
      "grad_norm": 0.6972466993378786,
      "learning_rate": 0.002661,
      "loss": 4.521,
      "step": 887
    },
    {
      "epoch": 0.00888,
      "grad_norm": 0.6171068656932988,
      "learning_rate": 0.002664,
      "loss": 4.5168,
      "step": 888
    },
    {
      "epoch": 0.00889,
      "grad_norm": 0.7028052332269225,
      "learning_rate": 0.002667,
      "loss": 4.4741,
      "step": 889
    },
    {
      "epoch": 0.0089,
      "grad_norm": 0.8226361692292167,
      "learning_rate": 0.00267,
      "loss": 4.538,
      "step": 890
    },
    {
      "epoch": 0.00891,
      "grad_norm": 1.1043997753739854,
      "learning_rate": 0.002673,
      "loss": 4.5318,
      "step": 891
    },
    {
      "epoch": 0.00892,
      "grad_norm": 0.9946974632170549,
      "learning_rate": 0.002676,
      "loss": 4.5463,
      "step": 892
    },
    {
      "epoch": 0.00893,
      "grad_norm": 0.8958086490440859,
      "learning_rate": 0.002679,
      "loss": 4.5224,
      "step": 893
    },
    {
      "epoch": 0.00894,
      "grad_norm": 0.9077012235708041,
      "learning_rate": 0.002682,
      "loss": 4.5134,
      "step": 894
    },
    {
      "epoch": 0.00895,
      "grad_norm": 0.8017398027780225,
      "learning_rate": 0.0026850000000000003,
      "loss": 4.5213,
      "step": 895
    },
    {
      "epoch": 0.00896,
      "grad_norm": 0.7045937615557313,
      "learning_rate": 0.0026880000000000003,
      "loss": 4.5092,
      "step": 896
    },
    {
      "epoch": 0.00897,
      "grad_norm": 0.6210872063297819,
      "learning_rate": 0.0026910000000000002,
      "loss": 4.5215,
      "step": 897
    },
    {
      "epoch": 0.00898,
      "grad_norm": 0.6465621097725943,
      "learning_rate": 0.002694,
      "loss": 4.5135,
      "step": 898
    },
    {
      "epoch": 0.00899,
      "grad_norm": 0.7170710385058783,
      "learning_rate": 0.002697,
      "loss": 4.5216,
      "step": 899
    },
    {
      "epoch": 0.009,
      "grad_norm": 0.6908740128767943,
      "learning_rate": 0.0027,
      "loss": 4.5448,
      "step": 900
    },
    {
      "epoch": 0.00901,
      "grad_norm": 0.6330297320675389,
      "learning_rate": 0.002703,
      "loss": 4.4948,
      "step": 901
    },
    {
      "epoch": 0.00902,
      "grad_norm": 0.6888151977010172,
      "learning_rate": 0.002706,
      "loss": 4.5079,
      "step": 902
    },
    {
      "epoch": 0.00903,
      "grad_norm": 0.7064881100348057,
      "learning_rate": 0.002709,
      "loss": 4.5094,
      "step": 903
    },
    {
      "epoch": 0.00904,
      "grad_norm": 0.6240828405072839,
      "learning_rate": 0.002712,
      "loss": 4.4821,
      "step": 904
    },
    {
      "epoch": 0.00905,
      "grad_norm": 0.5621159533559194,
      "learning_rate": 0.002715,
      "loss": 4.4943,
      "step": 905
    },
    {
      "epoch": 0.00906,
      "grad_norm": 0.6881074752193249,
      "learning_rate": 0.002718,
      "loss": 4.5117,
      "step": 906
    },
    {
      "epoch": 0.00907,
      "grad_norm": 0.8314783658771728,
      "learning_rate": 0.0027210000000000003,
      "loss": 4.5324,
      "step": 907
    },
    {
      "epoch": 0.00908,
      "grad_norm": 0.8326381841713656,
      "learning_rate": 0.0027240000000000003,
      "loss": 4.5097,
      "step": 908
    },
    {
      "epoch": 0.00909,
      "grad_norm": 0.9853385359973519,
      "learning_rate": 0.0027270000000000003,
      "loss": 4.536,
      "step": 909
    },
    {
      "epoch": 0.0091,
      "grad_norm": 1.1119575374996422,
      "learning_rate": 0.0027300000000000002,
      "loss": 4.5599,
      "step": 910
    },
    {
      "epoch": 0.00911,
      "grad_norm": 1.0405135652045494,
      "learning_rate": 0.002733,
      "loss": 4.5105,
      "step": 911
    },
    {
      "epoch": 0.00912,
      "grad_norm": 1.210969720326656,
      "learning_rate": 0.002736,
      "loss": 4.5363,
      "step": 912
    },
    {
      "epoch": 0.00913,
      "grad_norm": 0.9746238769615785,
      "learning_rate": 0.002739,
      "loss": 4.5155,
      "step": 913
    },
    {
      "epoch": 0.00914,
      "grad_norm": 1.1317971223126129,
      "learning_rate": 0.002742,
      "loss": 4.5705,
      "step": 914
    },
    {
      "epoch": 0.00915,
      "grad_norm": 0.8132490227755053,
      "learning_rate": 0.002745,
      "loss": 4.5092,
      "step": 915
    },
    {
      "epoch": 0.00916,
      "grad_norm": 1.1120180148538203,
      "learning_rate": 0.002748,
      "loss": 4.5111,
      "step": 916
    },
    {
      "epoch": 0.00917,
      "grad_norm": 1.0918567511234711,
      "learning_rate": 0.002751,
      "loss": 4.5277,
      "step": 917
    },
    {
      "epoch": 0.00918,
      "grad_norm": 0.9369545872786895,
      "learning_rate": 0.0027540000000000004,
      "loss": 4.5356,
      "step": 918
    },
    {
      "epoch": 0.00919,
      "grad_norm": 0.9217438795079191,
      "learning_rate": 0.0027570000000000003,
      "loss": 4.5218,
      "step": 919
    },
    {
      "epoch": 0.0092,
      "grad_norm": 0.862557634576406,
      "learning_rate": 0.0027600000000000003,
      "loss": 4.5305,
      "step": 920
    },
    {
      "epoch": 0.00921,
      "grad_norm": 1.1407056864076377,
      "learning_rate": 0.0027630000000000003,
      "loss": 4.533,
      "step": 921
    },
    {
      "epoch": 0.00922,
      "grad_norm": 1.0450100884646383,
      "learning_rate": 0.0027660000000000002,
      "loss": 4.5481,
      "step": 922
    },
    {
      "epoch": 0.00923,
      "grad_norm": 1.1216174249826327,
      "learning_rate": 0.002769,
      "loss": 4.5266,
      "step": 923
    },
    {
      "epoch": 0.00924,
      "grad_norm": 0.8159916660782444,
      "learning_rate": 0.002772,
      "loss": 4.534,
      "step": 924
    },
    {
      "epoch": 0.00925,
      "grad_norm": 0.8122347656711312,
      "learning_rate": 0.002775,
      "loss": 4.5145,
      "step": 925
    },
    {
      "epoch": 0.00926,
      "grad_norm": 1.0333326209619593,
      "learning_rate": 0.002778,
      "loss": 4.5431,
      "step": 926
    },
    {
      "epoch": 0.00927,
      "grad_norm": 1.028655155327321,
      "learning_rate": 0.002781,
      "loss": 4.509,
      "step": 927
    },
    {
      "epoch": 0.00928,
      "grad_norm": 0.9454951825640333,
      "learning_rate": 0.002784,
      "loss": 4.5533,
      "step": 928
    },
    {
      "epoch": 0.00929,
      "grad_norm": 1.1164441967376633,
      "learning_rate": 0.0027870000000000004,
      "loss": 4.5458,
      "step": 929
    },
    {
      "epoch": 0.0093,
      "grad_norm": 0.8645616643837938,
      "learning_rate": 0.0027900000000000004,
      "loss": 4.5148,
      "step": 930
    },
    {
      "epoch": 0.00931,
      "grad_norm": 0.9852728030510483,
      "learning_rate": 0.0027930000000000003,
      "loss": 4.5315,
      "step": 931
    },
    {
      "epoch": 0.00932,
      "grad_norm": 0.8717314730884365,
      "learning_rate": 0.0027960000000000003,
      "loss": 4.5342,
      "step": 932
    },
    {
      "epoch": 0.00933,
      "grad_norm": 0.9689543139825987,
      "learning_rate": 0.0027990000000000003,
      "loss": 4.5061,
      "step": 933
    },
    {
      "epoch": 0.00934,
      "grad_norm": 0.8011334760420642,
      "learning_rate": 0.0028020000000000002,
      "loss": 4.5228,
      "step": 934
    },
    {
      "epoch": 0.00935,
      "grad_norm": 0.7842131538830379,
      "learning_rate": 0.002805,
      "loss": 4.5333,
      "step": 935
    },
    {
      "epoch": 0.00936,
      "grad_norm": 0.8009313182730428,
      "learning_rate": 0.002808,
      "loss": 4.5328,
      "step": 936
    },
    {
      "epoch": 0.00937,
      "grad_norm": 0.6974240482478135,
      "learning_rate": 0.002811,
      "loss": 4.5386,
      "step": 937
    },
    {
      "epoch": 0.00938,
      "grad_norm": 0.5643540587859054,
      "learning_rate": 0.002814,
      "loss": 4.5083,
      "step": 938
    },
    {
      "epoch": 0.00939,
      "grad_norm": 0.5748911957324845,
      "learning_rate": 0.002817,
      "loss": 4.5002,
      "step": 939
    },
    {
      "epoch": 0.0094,
      "grad_norm": 0.5120628706209022,
      "learning_rate": 0.00282,
      "loss": 4.4948,
      "step": 940
    },
    {
      "epoch": 0.00941,
      "grad_norm": 0.4939803780583287,
      "learning_rate": 0.002823,
      "loss": 4.4711,
      "step": 941
    },
    {
      "epoch": 0.00942,
      "grad_norm": 0.5149843098575663,
      "learning_rate": 0.002826,
      "loss": 4.4896,
      "step": 942
    },
    {
      "epoch": 0.00943,
      "grad_norm": 0.5813616468237084,
      "learning_rate": 0.002829,
      "loss": 4.4876,
      "step": 943
    },
    {
      "epoch": 0.00944,
      "grad_norm": 0.6406019504044549,
      "learning_rate": 0.002832,
      "loss": 4.4895,
      "step": 944
    },
    {
      "epoch": 0.00945,
      "grad_norm": 0.5680854338022571,
      "learning_rate": 0.002835,
      "loss": 4.4913,
      "step": 945
    },
    {
      "epoch": 0.00946,
      "grad_norm": 0.5220321709660245,
      "learning_rate": 0.002838,
      "loss": 4.4951,
      "step": 946
    },
    {
      "epoch": 0.00947,
      "grad_norm": 0.631763941258451,
      "learning_rate": 0.0028409999999999998,
      "loss": 4.4705,
      "step": 947
    },
    {
      "epoch": 0.00948,
      "grad_norm": 0.6449801035821655,
      "learning_rate": 0.0028439999999999997,
      "loss": 4.4569,
      "step": 948
    },
    {
      "epoch": 0.00949,
      "grad_norm": 0.5961885502547112,
      "learning_rate": 0.002847,
      "loss": 4.4903,
      "step": 949
    },
    {
      "epoch": 0.0095,
      "grad_norm": 0.6511958288987238,
      "learning_rate": 0.00285,
      "loss": 4.4834,
      "step": 950
    },
    {
      "epoch": 0.00951,
      "grad_norm": 0.7136770979597694,
      "learning_rate": 0.002853,
      "loss": 4.4642,
      "step": 951
    },
    {
      "epoch": 0.00952,
      "grad_norm": 0.7943173544375755,
      "learning_rate": 0.002856,
      "loss": 4.4903,
      "step": 952
    },
    {
      "epoch": 0.00953,
      "grad_norm": 0.7165313674775572,
      "learning_rate": 0.002859,
      "loss": 4.4283,
      "step": 953
    },
    {
      "epoch": 0.00954,
      "grad_norm": 0.7629917982284083,
      "learning_rate": 0.002862,
      "loss": 4.49,
      "step": 954
    },
    {
      "epoch": 0.00955,
      "grad_norm": 0.9137416767731276,
      "learning_rate": 0.002865,
      "loss": 4.4809,
      "step": 955
    },
    {
      "epoch": 0.00956,
      "grad_norm": 0.9350410452513611,
      "learning_rate": 0.002868,
      "loss": 4.4685,
      "step": 956
    },
    {
      "epoch": 0.00957,
      "grad_norm": 0.8266195329744859,
      "learning_rate": 0.002871,
      "loss": 4.4866,
      "step": 957
    },
    {
      "epoch": 0.00958,
      "grad_norm": 0.7685667720087905,
      "learning_rate": 0.002874,
      "loss": 4.469,
      "step": 958
    },
    {
      "epoch": 0.00959,
      "grad_norm": 0.7114722439549117,
      "learning_rate": 0.002877,
      "loss": 4.462,
      "step": 959
    },
    {
      "epoch": 0.0096,
      "grad_norm": 0.7212359661166219,
      "learning_rate": 0.0028799999999999997,
      "loss": 4.4615,
      "step": 960
    },
    {
      "epoch": 0.00961,
      "grad_norm": 0.6952146282717937,
      "learning_rate": 0.002883,
      "loss": 4.4332,
      "step": 961
    },
    {
      "epoch": 0.00962,
      "grad_norm": 0.631327349964468,
      "learning_rate": 0.002886,
      "loss": 4.4607,
      "step": 962
    },
    {
      "epoch": 0.00963,
      "grad_norm": 0.7386203674854516,
      "learning_rate": 0.002889,
      "loss": 4.4627,
      "step": 963
    },
    {
      "epoch": 0.00964,
      "grad_norm": 0.7099703519669538,
      "learning_rate": 0.002892,
      "loss": 4.4494,
      "step": 964
    },
    {
      "epoch": 0.00965,
      "grad_norm": 0.6798689010174963,
      "learning_rate": 0.002895,
      "loss": 4.4243,
      "step": 965
    },
    {
      "epoch": 0.00966,
      "grad_norm": 0.7248321886037777,
      "learning_rate": 0.002898,
      "loss": 4.4416,
      "step": 966
    },
    {
      "epoch": 0.00967,
      "grad_norm": 0.6988539316842848,
      "learning_rate": 0.002901,
      "loss": 4.4803,
      "step": 967
    },
    {
      "epoch": 0.00968,
      "grad_norm": 0.7131963881573582,
      "learning_rate": 0.002904,
      "loss": 4.4431,
      "step": 968
    },
    {
      "epoch": 0.00969,
      "grad_norm": 0.6774288908313462,
      "learning_rate": 0.002907,
      "loss": 4.4758,
      "step": 969
    },
    {
      "epoch": 0.0097,
      "grad_norm": 0.7070891862905684,
      "learning_rate": 0.00291,
      "loss": 4.4353,
      "step": 970
    },
    {
      "epoch": 0.00971,
      "grad_norm": 0.5977635294593704,
      "learning_rate": 0.002913,
      "loss": 4.4375,
      "step": 971
    },
    {
      "epoch": 0.00972,
      "grad_norm": 0.6000392249111391,
      "learning_rate": 0.002916,
      "loss": 4.4443,
      "step": 972
    },
    {
      "epoch": 0.00973,
      "grad_norm": 0.5652705565921429,
      "learning_rate": 0.002919,
      "loss": 4.4351,
      "step": 973
    },
    {
      "epoch": 0.00974,
      "grad_norm": 0.5862941407898875,
      "learning_rate": 0.002922,
      "loss": 4.4242,
      "step": 974
    },
    {
      "epoch": 0.00975,
      "grad_norm": 0.5761106589533092,
      "learning_rate": 0.002925,
      "loss": 4.4454,
      "step": 975
    },
    {
      "epoch": 0.00976,
      "grad_norm": 0.6267951053667676,
      "learning_rate": 0.002928,
      "loss": 4.4304,
      "step": 976
    },
    {
      "epoch": 0.00977,
      "grad_norm": 0.7323160890025502,
      "learning_rate": 0.002931,
      "loss": 4.4742,
      "step": 977
    },
    {
      "epoch": 0.00978,
      "grad_norm": 0.7947149437193983,
      "learning_rate": 0.002934,
      "loss": 4.4506,
      "step": 978
    },
    {
      "epoch": 0.00979,
      "grad_norm": 0.7190657481675354,
      "learning_rate": 0.002937,
      "loss": 4.4415,
      "step": 979
    },
    {
      "epoch": 0.0098,
      "grad_norm": 0.7127444247767541,
      "learning_rate": 0.00294,
      "loss": 4.4341,
      "step": 980
    },
    {
      "epoch": 0.00981,
      "grad_norm": 0.6752689834635909,
      "learning_rate": 0.002943,
      "loss": 4.4448,
      "step": 981
    },
    {
      "epoch": 0.00982,
      "grad_norm": 0.7621208928943479,
      "learning_rate": 0.002946,
      "loss": 4.4402,
      "step": 982
    },
    {
      "epoch": 0.00983,
      "grad_norm": 0.6278907346255873,
      "learning_rate": 0.0029490000000000002,
      "loss": 4.4356,
      "step": 983
    },
    {
      "epoch": 0.00984,
      "grad_norm": 0.640256203275423,
      "learning_rate": 0.002952,
      "loss": 4.4238,
      "step": 984
    },
    {
      "epoch": 0.00985,
      "grad_norm": 0.6571839667979233,
      "learning_rate": 0.002955,
      "loss": 4.4427,
      "step": 985
    },
    {
      "epoch": 0.00986,
      "grad_norm": 0.8077439009351965,
      "learning_rate": 0.002958,
      "loss": 4.4152,
      "step": 986
    },
    {
      "epoch": 0.00987,
      "grad_norm": 0.8060461488431017,
      "learning_rate": 0.002961,
      "loss": 4.4298,
      "step": 987
    },
    {
      "epoch": 0.00988,
      "grad_norm": 0.8493960405917234,
      "learning_rate": 0.002964,
      "loss": 4.4644,
      "step": 988
    },
    {
      "epoch": 0.00989,
      "grad_norm": 0.7333952379065988,
      "learning_rate": 0.002967,
      "loss": 4.405,
      "step": 989
    },
    {
      "epoch": 0.0099,
      "grad_norm": 0.6561442246030965,
      "learning_rate": 0.00297,
      "loss": 4.4391,
      "step": 990
    },
    {
      "epoch": 0.00991,
      "grad_norm": 0.760469930109403,
      "learning_rate": 0.002973,
      "loss": 4.4396,
      "step": 991
    },
    {
      "epoch": 0.00992,
      "grad_norm": 0.7424003895138885,
      "learning_rate": 0.002976,
      "loss": 4.4355,
      "step": 992
    },
    {
      "epoch": 0.00993,
      "grad_norm": 0.7430266355026064,
      "learning_rate": 0.002979,
      "loss": 4.4325,
      "step": 993
    },
    {
      "epoch": 0.00994,
      "grad_norm": 0.7650486142940663,
      "learning_rate": 0.002982,
      "loss": 4.4406,
      "step": 994
    },
    {
      "epoch": 0.00995,
      "grad_norm": 0.7876492554203988,
      "learning_rate": 0.0029850000000000002,
      "loss": 4.4395,
      "step": 995
    },
    {
      "epoch": 0.00996,
      "grad_norm": 0.7683559647577115,
      "learning_rate": 0.002988,
      "loss": 4.4448,
      "step": 996
    },
    {
      "epoch": 0.00997,
      "grad_norm": 0.7326948857624186,
      "learning_rate": 0.002991,
      "loss": 4.432,
      "step": 997
    },
    {
      "epoch": 0.00998,
      "grad_norm": 0.8133968945705032,
      "learning_rate": 0.002994,
      "loss": 4.4447,
      "step": 998
    },
    {
      "epoch": 0.00999,
      "grad_norm": 0.9410152737222023,
      "learning_rate": 0.002997,
      "loss": 4.4373,
      "step": 999
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.799133553847171,
      "learning_rate": 0.003,
      "loss": 4.4596,
      "step": 1000
    }
  ],
  "logging_steps": 1,
  "max_steps": 100000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 25,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.9643642855424e+16,
  "train_batch_size": 1024,
  "trial_name": null,
  "trial_params": null
}